Updated the regression dataset for pr-300

Signed-off-by: Peter Staar <taa@zurich.ibm.com>

groundtruth/parser/complex_invisible_fonts_04.pdf.page_no_1.py.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:057698b0ba821f0250c34468e80da0ebf65f815560638704ec65fc3110f6de7c
-size 2284569
+oid sha256:4e97ebf826f0caa13a4e74b730c98343f8e5372679a4108eb52e0ad47fffb3f6
+size 2285241

groundtruth/parser/complex_invisible_fonts_04.pdf.page_no_1.py.json.char.txt

@@ -251,7 +251,7 @@
 (292.62, 479.47) (296.74, 479.47) (296.74, 486.05) (292.62, 486.05)      /T1_0 u	<|special_separator|>
 (296.89, 479.47) (299.78, 479.47) (299.78, 486.05) (296.89, 486.05)      /T1_0 t	<|special_separator|>
 (299.92, 479.47) (304.15, 479.47) (304.15, 486.05) (299.92, 486.05)      /T1_0 h	<|special_separator|>
-(304.30, 479.47) (306.01, 479.47) (306.01, 486.05) (304.30, 486.05)      /T1_0  	<|special_separator|>
+(304.30, 479.47) (306.01, 479.47) (306.01, 486.05) (304.30, 486.05)      /T1_0  	<|special_separator|>
 (306.16, 479.47) (311.22, 479.47) (311.22, 486.05) (306.16, 486.05)      /T1_0 A	<|special_separator|>
 (311.37, 479.47) (313.97, 479.47) (313.97, 486.05) (311.37, 486.05)      /T1_0 f	<|special_separator|>
 (314.11, 479.47) (316.48, 479.47) (316.48, 486.05) (314.11, 486.05)      /T1_0 r	<|special_separator|>
@@ -446,7 +446,7 @@
 (491.91, 718.32) (495.09, 718.32) (495.09, 725.57) (491.91, 725.57)      /T1_0 t	<|special_separator|>
 (495.25, 718.32) (498.91, 718.32) (498.91, 725.57) (495.25, 725.57)      /T1_0 s	<|special_separator|>
 (498.91, 718.32) (500.80, 718.32) (500.80, 725.57) (498.91, 725.57)      /T1_0  	<|special_separator|>
-(042.52, 705.32) (044.41, 705.32) (044.41, 712.57) (042.52, 712.57)      /T1_0  	<|special_separator|>
+(042.52, 705.32) (044.41, 705.32) (044.41, 712.57) (042.52, 712.57)      /T1_0 ­	<|special_separator|>
 (042.52, 705.32) (046.62, 705.32) (046.62, 712.57) (042.52, 712.57)      /T1_0 c	<|special_separator|>
 (046.78, 705.32) (051.46, 705.32) (051.46, 712.57) (046.78, 712.57)      /T1_0 o	<|special_separator|>
 (051.62, 705.32) (058.83, 705.32) (058.83, 712.57) (051.62, 712.57)      /T1_0 m	<|special_separator|>

groundtruth/parser/complex_invisible_fonts_04.pdf.page_no_1.py.json.line.txt

@@ -7,11 +7,11 @@
 (116.08, 479.47) (146.70, 479.47) (146.70, 486.05) (116.08, 486.05)      /T1_0 Germany	<|special_separator|>
 (210.79, 479.47) (227.74, 479.47) (227.74, 486.05) (210.79, 486.05)      /T1_0 India	<|special_separator|>
 (246.84, 479.47) (276.72, 479.47) (276.72, 486.05) (246.84, 486.05)      /T1_0 Pakistan	<|special_separator|>
-(283.70, 479.47) (326.60, 479.47) (326.60, 486.05) (283.70, 486.05)      /T1_0 South Africa	<|special_separator|>
+(283.70, 479.47) (326.60, 479.47) (326.60, 486.05) (283.70, 486.05)      /T1_0 South Africa	<|special_separator|>
 (332.83, 479.47) (361.38, 479.47) (361.38, 486.05) (332.83, 486.05)      /T1_0 Ethiopia	<|special_separator|>
 (042.60, 743.54) (500.94, 743.54) (500.94, 755.25) (042.60, 755.25)      /T1_1 SAME PRODUCT, DIFFERENT MARKETS: OBVIOUS DOUBLE STANDARD	<|special_separator|>
 (042.52, 718.32) (500.80, 718.32) (500.80, 725.57) (042.52, 725.57)      /T1_0 Wheat-based infant cereals with milk are Nestlé's founding product. Today, Cerelac is sold all over the world, although its 	<|special_separator|>
-(042.52, 705.32) (258.21, 705.32) (258.21, 712.57) (042.52, 712.57)      /T1_0   composition varies considerably from country to country.	<|special_separator|>
+(042.52, 705.32) (258.21, 705.32) (258.21, 712.57) (042.52, 712.57)      /T1_0 ­ composition varies considerably from country to country.	<|special_separator|>
 (042.52, 655.58) (445.63, 655.58) (445.63, 664.58) (042.52, 664.58)      /T1_1 Added sugar content of the same Cerelac wheat product in different countries	<|special_separator|>
 (374.15, 479.47) (404.53, 479.47) (404.53, 486.05) (374.15, 486.05)      /T1_0 Thailand	<|special_separator|>
 (042.52, 314.80) (285.98, 314.80) (285.98, 323.69) (042.52, 323.69)      /T1_2 using idealizing imagery, it's  not  transparent  when  it 	<|special_separator|>

groundtruth/parser/complex_invisible_fonts_04.pdf.page_no_1.py.json.word.txt

@@ -64,6 +64,7 @@
 (428.85, 718.32) (451.25, 718.32) (451.25, 725.57) (428.85, 725.57)      /T1_0 world,	<|special_separator|>
 (453.45, 718.32) (487.64, 718.32) (487.64, 725.57) (453.45, 725.57)      /T1_0 although	<|special_separator|>
 (489.85, 718.32) (498.91, 718.32) (498.91, 725.57) (489.85, 725.57)      /T1_0 its	<|special_separator|>
+(042.52, 705.32) (044.41, 705.32) (044.41, 712.57) (042.52, 712.57)      /T1_0 ­	<|special_separator|>
 (042.52, 705.32) (089.57, 705.32) (089.57, 712.57) (042.52, 712.57)      /T1_0 composition	<|special_separator|>
 (091.78, 705.32) (113.86, 705.32) (113.86, 712.57) (091.78, 712.57)      /T1_0 varies	<|special_separator|>
 (116.06, 705.32) (164.03, 705.32) (164.03, 712.57) (116.06, 712.57)      /T1_0 considerably	<|special_separator|>

groundtruth/parser/complex_invisible_fonts_05.pdf.page_no_1.py.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f6aad9414ac69fe85ea5950108d9cf7b3827a474f9aeedb8dcf7cd75c81c4dee
-size 2715302
+oid sha256:805f748704ad53b33b2d50ceab139e5b1069f598e15ae4730e266b0dae57fb62
+size 2715959

groundtruth/parser/complex_invisible_fonts_05.pdf.page_no_1.py.json.char.txt

@@ -2062,7 +2062,7 @@
 (249.41, 054.80) (252.42, 054.80) (252.42, 063.69) (249.41, 063.69)      /T1_2 i	<|special_separator|>
 (252.47, 054.80) (258.47, 054.80) (258.47, 063.69) (252.47, 063.69)      /T1_2 n	<|special_separator|>
 (258.52, 054.80) (260.96, 054.80) (260.96, 063.69) (258.52, 063.69)      /T1_2  	<|special_separator|>
-(261.98, 054.80) (264.42, 054.80) (264.42, 063.69) (261.98, 063.69)      /T1_2  	<|special_separator|>
+(261.98, 054.80) (264.42, 054.80) (264.42, 063.69) (261.98, 063.69)      /T1_2 ­	<|special_separator|>
 (261.97, 054.80) (270.76, 054.80) (270.76, 063.69) (261.97, 063.69)      /T1_2 m	<|special_separator|>
 (270.81, 054.80) (275.51, 054.80) (275.51, 063.69) (270.81, 063.69)      /T1_2 a	<|special_separator|>
 (275.56, 054.80) (279.65, 054.80) (279.65, 063.69) (275.56, 063.69)      /T1_2 r	<|special_separator|>

groundtruth/parser/complex_invisible_fonts_05.pdf.page_no_1.py.json.line.txt

@@ -81,7 +81,7 @@
 (056.69, 093.80) (300.05, 093.80) (300.05, 102.69) (056.69, 102.69)      /T1_2 milks market. In 2022, global sales of Nido products for 	<|special_separator|>
 (056.69, 080.80) (297.64, 080.80) (297.64, 089.69) (056.69, 089.69)      /T1_2 young children aged one to three had exceeded $1 bil-	<|special_separator|>
 (056.69, 067.80) (300.13, 067.80) (300.13, 076.69) (056.69, 076.69)      /T1_2 lion according to Euromonitor. We examined 29 of these 	<|special_separator|>
-(056.69, 054.80) (300.09, 054.80) (300.09, 063.69) (056.69, 063.69)      /T1_2 products sold by Nestlé in some of the main   markets 	<|special_separator|>
+(056.69, 054.80) (300.09, 054.80) (300.09, 063.69) (056.69, 063.69)      /T1_2 products sold by Nestlé in some of the main ­ markets 	<|special_separator|>
 (311.81, 314.80) (555.21, 314.80) (555.21, 323.69) (311.81, 323.69)      /T1_2 in low- and middle-income countries. The result: 21 of 	<|special_separator|>
 (311.81, 301.80) (484.41, 301.80) (484.41, 310.69) (311.81, 310.69)      /T1_2 them (72 percent) contain added sugar. 	<|special_separator|>
 (340.15, 288.80) (552.76, 288.80) (552.76, 297.69) (340.15, 297.69)      /T1_2 For nine of these products, we were able to de-	<|special_separator|>

groundtruth/parser/complex_invisible_fonts_05.pdf.page_no_1.py.json.word.txt

@@ -371,6 +371,7 @@
 (205.74, 054.80) (214.71, 054.80) (214.71, 063.69) (205.74, 063.69)      /T1_2 of	<|special_separator|>
 (218.22, 054.80) (232.30, 054.80) (232.30, 063.69) (218.22, 063.69)      /T1_2 the	<|special_separator|>
 (235.81, 054.80) (258.47, 054.80) (258.47, 063.69) (235.81, 063.69)      /T1_2 main	<|special_separator|>
+(261.98, 054.80) (264.42, 054.80) (264.42, 063.69) (261.98, 063.69)      /T1_2 ­	<|special_separator|>
 (261.97, 054.80) (297.64, 054.80) (297.64, 063.69) (261.97, 063.69)      /T1_2 markets	<|special_separator|>
 (311.81, 314.80) (320.86, 314.80) (320.86, 323.69) (311.81, 323.69)      /T1_2 in	<|special_separator|>
 (323.56, 314.80) (343.90, 314.80) (343.90, 323.69) (323.56, 323.69)      /T1_2 low-	<|special_separator|>

groundtruth/parser/cropbox_versus_mediabox_01.pdf.page_no_1.py.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0864f134834a82d55a2bf36320ee6cebd55bd8b2a10e69a23d9653f5ac61c996
-size 3597312
+oid sha256:6a189f6baa28e837bc328fcc29c7e5c8dfe4c2e49f448106f012ca02ae9c78c8
+size 3604232

groundtruth/parser/cropbox_versus_mediabox_01.pdf.page_no_1.py.json.char.txt

@@ -17,7 +17,7 @@
 (058.27, 166.09) (068.27, 166.09) (068.27, 176.89) (058.27, 176.89)      /C0_1 ら	<|special_separator|>
 (068.27, 166.09) (078.27, 166.09) (078.27, 176.89) (068.27, 176.89)      /C0_1 せ	<|special_separator|>
 (034.02, 152.53) (040.32, 152.53) (040.32, 159.23) (034.02, 159.23)      /C0_2 ●	<|special_separator|>
-(040.32, 150.87) (043.31, 150.87) (043.31, 160.45) (040.32, 160.45)      /C0_2 	<|special_separator|>
+(040.32, 150.87) (043.31, 150.87) (043.31, 160.45) (040.32, 160.45)      /C0_2 		<|special_separator|>
 (043.09, 150.87) (052.09, 150.87) (052.09, 160.45) (043.09, 160.45)      /C0_2 ご	<|special_separator|>
 (052.09, 150.87) (061.09, 150.87) (061.09, 160.45) (052.09, 160.45)      /C0_2 購	<|special_separator|>
 (061.09, 150.87) (070.09, 150.87) (070.09, 160.45) (061.09, 160.45)      /C0_2 入	<|special_separator|>
@@ -38,7 +38,7 @@
 (070.99, 138.37) (079.99, 138.37) (079.99, 147.95) (070.99, 147.95)      /C0_2 す	<|special_separator|>
 (079.99, 138.37) (088.99, 138.37) (088.99, 147.95) (079.99, 147.95)      /C0_2 。	<|special_separator|>
 (034.02, 127.53) (040.32, 127.53) (040.32, 134.23) (034.02, 134.23)      /C0_2 ●	<|special_separator|>
-(040.32, 125.87) (043.31, 125.87) (043.31, 135.45) (040.32, 135.45)      /C0_2 	<|special_separator|>
+(040.32, 125.87) (043.31, 125.87) (043.31, 135.45) (040.32, 135.45)      /C0_2 		<|special_separator|>
 (043.02, 125.87) (052.02, 125.87) (052.02, 135.45) (043.02, 135.45)      /C0_2 自	<|special_separator|>
 (051.95, 125.87) (060.95, 125.87) (060.95, 135.45) (051.95, 135.45)      /C0_2 動	<|special_separator|>
 (060.89, 125.87) (069.89, 125.87) (069.89, 135.45) (060.89, 135.45)      /C0_2 内	<|special_separator|>
@@ -132,8 +132,8 @@
 (086.21, 047.50) (095.21, 047.50) (095.21, 057.08) (086.21, 057.08)      /C0_2 う	<|special_separator|>
 (094.80, 047.50) (103.80, 047.50) (103.80, 057.08) (094.80, 057.08)      /C0_2 一	<|special_separator|>
 (103.64, 047.50) (112.64, 047.50) (112.64, 057.08) (103.64, 057.08)      /C0_2 度	<|special_separator|>
-(112.53, 047.50) (115.53, 047.50) (115.53, 057.08) (112.53, 057.08)      /C0_2 	<|special_separator|>
-(146.47, 047.50) (149.47, 047.50) (149.47, 057.07) (146.47, 057.07)      /C0_2 	<|special_separator|>
+(112.53, 047.50) (115.53, 047.50) (115.53, 057.08) (112.53, 057.08)      /C0_2  	<|special_separator|>
+(146.47, 047.50) (149.47, 047.50) (149.47, 057.07) (146.47, 057.07)      /C0_2  	<|special_separator|>
 (149.29, 047.50) (158.29, 047.50) (158.29, 057.07) (149.29, 057.07)      /C0_2 を	<|special_separator|>
 (043.02, 032.98) (052.02, 032.98) (052.02, 042.56) (043.02, 042.56)      /C0_2 押	<|special_separator|>
 (051.38, 032.98) (060.38, 032.98) (060.38, 042.56) (051.38, 042.56)      /C0_2 し	<|special_separator|>
@@ -197,7 +197,7 @@
 (268.33, 174.79) (279.33, 174.79) (279.33, 186.62) (268.33, 186.62)      /C0_4 と	<|special_separator|>
 (279.33, 174.79) (290.33, 174.79) (290.33, 186.62) (279.33, 186.62)      /C0_4 は	<|special_separator|>
 (188.59, 163.26) (195.59, 163.26) (195.59, 170.70) (188.59, 170.70)      /C0_2 ●	<|special_separator|>
-(195.59, 161.41) (198.92, 161.41) (198.92, 172.05) (195.59, 172.05)      /C0_2 	<|special_separator|>
+(195.59, 161.41) (198.92, 161.41) (198.92, 172.05) (195.59, 172.05)      /C0_2 		<|special_separator|>
 (198.51, 161.41) (208.51, 161.41) (208.51, 172.05) (198.51, 172.05)      /C0_2 壁	<|special_separator|>
 (208.51, 161.41) (218.51, 161.41) (218.51, 172.05) (208.51, 172.05)      /C0_2 温	<|special_separator|>
 (218.51, 161.41) (228.51, 161.41) (228.51, 172.05) (218.51, 172.05)      /C0_2 度	<|special_separator|>
@@ -279,7 +279,7 @@
 (368.51, 137.41) (378.51, 137.41) (378.51, 148.05) (368.51, 148.05)      /C0_2 す	<|special_separator|>
 (378.51, 137.41) (388.51, 137.41) (388.51, 148.05) (378.51, 148.05)      /C0_2 。	<|special_separator|>
 (188.59, 124.42) (195.59, 124.42) (195.59, 131.87) (188.59, 131.87)      /C0_2 ●	<|special_separator|>
-(195.59, 122.58) (198.92, 122.58) (198.92, 133.22) (195.59, 133.22)      /C0_2 	<|special_separator|>
+(195.59, 122.58) (198.92, 122.58) (198.92, 133.22) (195.59, 133.22)      /C0_2 		<|special_separator|>
 (198.51, 122.58) (208.51, 122.58) (208.51, 133.22) (198.51, 133.22)      /C0_2 お	<|special_separator|>
 (208.51, 122.58) (218.51, 122.58) (218.51, 133.22) (208.51, 133.22)      /C0_2 好	<|special_separator|>
 (218.51, 122.58) (228.51, 122.58) (228.51, 133.22) (218.51, 133.22)      /C0_2 み	<|special_separator|>
@@ -396,13 +396,13 @@
 (249.00, 436.94) (265.00, 436.94) (265.00, 453.61) (249.00, 453.61)      /C0_5 い	<|special_separator|>
 (265.00, 436.94) (281.00, 436.94) (281.00, 453.61) (265.00, 453.61)      /C0_5 と	<|special_separator|>
 (281.00, 436.94) (297.00, 436.94) (297.00, 453.61) (281.00, 453.61)      /C0_5 き	<|special_separator|>
-(226.27, 411.01) (231.26, 411.01) (231.26, 427.21) (226.27, 427.21)      /C0_1 		<|special_separator|>
+(226.27, 411.01) (231.26, 411.01) (231.26, 427.21) (226.27, 427.21)      /C0_1  	<|special_separator|>
 (231.26, 411.01) (246.26, 411.01) (246.26, 427.21) (231.26, 427.21)      /C0_1 を	<|special_separator|>
 (246.26, 411.01) (261.26, 411.01) (261.26, 427.21) (246.26, 427.21)      /C0_1 押	<|special_separator|>
 (261.26, 411.01) (276.26, 411.01) (276.26, 427.21) (261.26, 427.21)      /C0_1 す	<|special_separator|>
 (276.26, 411.01) (291.26, 411.01) (291.26, 427.21) (276.26, 427.21)      /C0_1 。	<|special_separator|>
 (185.00, 393.97) (192.00, 393.97) (192.00, 401.42) (185.00, 401.42)      /C0_2 ●	<|special_separator|>
-(192.00, 392.13) (195.33, 392.13) (195.33, 402.77) (192.00, 402.77)      /C0_2 	<|special_separator|>
+(192.00, 392.13) (195.33, 392.13) (195.33, 402.77) (192.00, 402.77)      /C0_2 		<|special_separator|>
 (189.92, 392.13) (199.92, 392.13) (199.92, 402.77) (189.92, 402.77)      /C0_2 「	<|special_separator|>
 (199.92, 392.13) (209.92, 392.13) (209.92, 402.77) (199.92, 402.77)      /C0_2 停	<|special_separator|>
 (209.92, 392.13) (219.92, 392.13) (219.92, 402.77) (209.92, 402.77)      /C0_2 止	<|special_separator|>
@@ -428,7 +428,7 @@
 (204.92, 368.13) (214.92, 368.13) (214.92, 378.77) (204.92, 378.77)      /C0_2 す	<|special_separator|>
 (214.92, 368.13) (224.92, 368.13) (224.92, 378.77) (214.92, 378.77)      /C0_2 。	<|special_separator|>
 (185.00, 354.30) (192.00, 354.30) (192.00, 361.75) (185.00, 361.75)      /C0_2 ●	<|special_separator|>
-(192.00, 352.46) (195.33, 352.46) (195.33, 363.10) (192.00, 363.10)      /C0_2 	<|special_separator|>
+(192.00, 352.46) (195.33, 352.46) (195.33, 363.10) (192.00, 363.10)      /C0_2 		<|special_separator|>
 (194.92, 352.46) (204.92, 352.46) (204.92, 363.10) (194.92, 363.10)      /C0_2 自	<|special_separator|>
 (204.92, 352.46) (214.92, 352.46) (214.92, 363.10) (204.92, 363.10)      /C0_2 動	<|special_separator|>
 (214.92, 352.46) (224.92, 352.46) (224.92, 363.10) (214.92, 363.10)      /C0_2 内	<|special_separator|>
@@ -451,8 +451,8 @@
 (204.92, 338.94) (214.92, 338.94) (214.92, 349.58) (204.92, 349.58)      /C0_2 う	<|special_separator|>
 (214.92, 338.94) (224.92, 338.94) (224.92, 349.58) (214.92, 349.58)      /C0_2 一	<|special_separator|>
 (224.92, 338.94) (234.92, 338.94) (234.92, 349.58) (224.92, 349.58)      /C0_2 度	<|special_separator|>
-(234.92, 338.94) (238.25, 338.94) (238.25, 349.58) (234.92, 349.58)      /C0_2 	<|special_separator|>
-(269.20, 338.94) (272.53, 338.94) (272.53, 349.58) (269.20, 349.58)      /C0_2 	<|special_separator|>
+(234.92, 338.94) (238.25, 338.94) (238.25, 349.58) (234.92, 349.58)      /C0_2  	<|special_separator|>
+(269.20, 338.94) (272.53, 338.94) (272.53, 349.58) (269.20, 349.58)      /C0_2  	<|special_separator|>
 (272.53, 338.94) (282.53, 338.94) (282.53, 349.58) (272.53, 349.58)      /C0_2 を	<|special_separator|>
 (282.53, 338.94) (292.53, 338.94) (292.53, 349.58) (282.53, 349.58)      /C0_2 押	<|special_separator|>
 (292.53, 338.94) (302.53, 338.94) (302.53, 349.58) (292.53, 349.58)      /C0_2 し	<|special_separator|>
@@ -491,13 +491,13 @@
 (313.00, 522.47) (329.00, 522.47) (329.00, 539.14) (313.00, 539.14)      /C0_5 い	<|special_separator|>
 (329.00, 522.47) (345.00, 522.47) (345.00, 539.14) (329.00, 539.14)      /C0_5 と	<|special_separator|>
 (345.00, 522.47) (361.00, 522.47) (361.00, 539.14) (345.00, 539.14)      /C0_5 き	<|special_separator|>
-(208.54, 486.42) (213.54, 486.42) (213.54, 502.62) (208.54, 502.62)      /C0_1 		<|special_separator|>
+(208.54, 486.42) (213.54, 486.42) (213.54, 502.62) (208.54, 502.62)      /C0_1  	<|special_separator|>
 (213.54, 486.42) (228.54, 486.42) (228.54, 502.62) (213.54, 502.62)      /C0_1 を	<|special_separator|>
 (228.54, 486.42) (243.54, 486.42) (243.54, 502.62) (228.54, 502.62)      /C0_1 押	<|special_separator|>
 (243.54, 486.42) (258.54, 486.42) (258.54, 502.62) (243.54, 502.62)      /C0_1 す	<|special_separator|>
 (258.54, 486.42) (273.54, 486.42) (273.54, 502.62) (258.54, 502.62)      /C0_1 。	<|special_separator|>
 (284.21, 506.09) (291.21, 506.09) (291.21, 513.54) (284.21, 513.54)      /C0_2 ●	<|special_separator|>
-(291.21, 504.25) (294.54, 504.25) (294.54, 514.89) (291.21, 514.89)      /C0_2 	<|special_separator|>
+(291.21, 504.25) (294.54, 504.25) (294.54, 514.89) (291.21, 514.89)      /C0_2 		<|special_separator|>
 (294.13, 504.25) (304.13, 504.25) (304.13, 514.89) (294.13, 514.89)      /C0_2 適	<|special_separator|>
 (304.13, 504.25) (314.13, 504.25) (314.13, 514.89) (304.13, 514.89)      /C0_2 温	<|special_separator|>
 (314.14, 510.14) (319.97, 510.14) (319.97, 516.34) (314.14, 516.34)      /C0_2 ※	<|special_separator|>
@@ -509,9 +509,9 @@
 (366.08, 504.25) (369.04, 504.25) (369.04, 514.89) (366.08, 514.89)      /C0_2 .	<|special_separator|>
 (369.04, 504.25) (375.15, 504.25) (375.15, 514.89) (369.04, 514.89)      /C0_2 0	<|special_separator|>
 (375.15, 504.25) (385.15, 504.25) (385.15, 514.89) (375.15, 514.89)      /C0_2 ℃	<|special_separator|>
-(385.15, 504.25) (388.48, 504.25) (388.48, 514.89) (385.15, 514.89)      /C0_2 	<|special_separator|>
+(385.15, 504.25) (388.48, 504.25) (388.48, 514.89) (385.15, 514.89)      /C0_2  	<|special_separator|>
 (388.48, 504.25) (396.63, 504.25) (396.63, 514.89) (388.48, 514.89)      /C0_2 ~	<|special_separator|>
-(396.63, 504.25) (399.96, 504.25) (399.96, 514.89) (396.63, 514.89)      /C0_2 	<|special_separator|>
+(396.63, 504.25) (399.96, 504.25) (399.96, 514.89) (396.63, 514.89)      /C0_2  	<|special_separator|>
 (399.96, 504.25) (409.96, 504.25) (409.96, 514.89) (399.96, 514.89)      /C0_2 ＋	<|special_separator|>
 (409.96, 504.25) (416.07, 504.25) (416.07, 514.89) (409.96, 514.89)      /C0_2 5	<|special_separator|>
 (416.07, 504.25) (419.03, 504.25) (419.03, 514.89) (416.07, 514.89)      /C0_2 .	<|special_separator|>
@@ -540,7 +540,7 @@
 (354.32, 492.25) (364.32, 492.25) (364.32, 502.89) (354.32, 502.89)      /C0_2 ）	<|special_separator|>
 (359.32, 492.25) (362.65, 492.25) (362.65, 502.89) (359.32, 502.89)      /C0_2 	<|special_separator|>
 (293.76, 480.25) (302.36, 480.25) (302.36, 490.89) (293.76, 490.89)      /C0_2 ※	<|special_separator|>
-(301.91, 480.25) (304.77, 480.25) (304.77, 490.89) (301.91, 490.89)      /C0_2 	<|special_separator|>
+(301.91, 480.25) (304.77, 480.25) (304.77, 490.89) (301.91, 490.89)      /C0_2 	<|special_separator|>
 (301.82, 480.25) (310.42, 480.25) (310.42, 490.89) (301.82, 490.89)      /C0_2 適	<|special_separator|>
 (310.16, 480.25) (318.76, 480.25) (318.76, 490.89) (310.16, 490.89)      /C0_2 温	<|special_separator|>
 (318.33, 480.25) (326.93, 480.25) (326.93, 490.89) (318.33, 490.89)      /C0_2 と	<|special_separator|>
@@ -585,7 +585,7 @@
 (345.00, 301.98) (361.00, 301.98) (361.00, 318.65) (345.00, 318.65)      /C0_5 と	<|special_separator|>
 (361.00, 301.98) (377.00, 301.98) (377.00, 318.65) (361.00, 318.65)      /C0_5 き	<|special_separator|>
 (185.00, 284.90) (192.00, 284.90) (192.00, 292.34) (185.00, 292.34)      /C0_2 ●	<|special_separator|>
-(192.00, 283.05) (195.33, 283.05) (195.33, 293.69) (192.00, 293.69)      /C0_2 	<|special_separator|>
+(192.00, 283.05) (195.33, 283.05) (195.33, 293.69) (192.00, 293.69)      /C0_2 		<|special_separator|>
 (194.92, 283.05) (204.92, 283.05) (204.92, 293.69) (194.92, 293.69)      /C0_2 エ	<|special_separator|>
 (204.92, 283.05) (214.92, 283.05) (214.92, 293.69) (204.92, 293.69)      /C0_2 ア	<|special_separator|>
 (214.92, 283.05) (224.92, 283.05) (224.92, 293.69) (214.92, 293.69)      /C0_2 コ	<|special_separator|>
@@ -609,7 +609,7 @@
 (234.92, 271.05) (244.92, 271.05) (244.92, 281.69) (234.92, 281.69)      /C0_2 は	<|special_separator|>
 (244.92, 271.05) (254.92, 271.05) (254.92, 281.69) (244.92, 281.69)      /C0_2 、	<|special_separator|>
 (249.92, 271.05) (253.25, 271.05) (253.25, 281.69) (249.92, 281.69)      /C0_2 	<|special_separator|>
-(227.41, 255.04) (230.74, 255.04) (230.74, 265.68) (227.41, 265.68)      /C0_2 	<|special_separator|>
+(227.41, 255.04) (230.74, 255.04) (230.74, 265.68) (227.41, 265.68)      /C0_2  	<|special_separator|>
 (228.66, 255.04) (238.66, 255.04) (238.66, 265.68) (228.66, 265.68)      /C0_2 で	<|special_separator|>
 (238.66, 255.04) (248.66, 255.04) (248.66, 265.68) (238.66, 265.68)      /C0_2 確	<|special_separator|>
 (248.66, 255.04) (258.66, 255.04) (258.66, 265.68) (248.66, 265.68)      /C0_2 認	<|special_separator|>
@@ -619,7 +619,7 @@
 (288.66, 255.04) (298.66, 255.04) (298.66, 265.68) (288.66, 265.68)      /C0_2 す	<|special_separator|>
 (298.66, 255.04) (308.66, 255.04) (308.66, 265.68) (298.66, 265.68)      /C0_2 。	<|special_separator|>
 (303.66, 255.04) (306.99, 255.04) (306.99, 265.68) (303.66, 265.68)      /C0_2 	<|special_separator|>
-(227.41, 235.01) (230.74, 235.01) (230.74, 245.64) (227.41, 245.64)      /C0_2 	<|special_separator|>
+(227.41, 235.01) (230.74, 235.01) (230.74, 245.64) (227.41, 245.64)      /C0_2  	<|special_separator|>
 (228.66, 235.01) (238.66, 235.01) (238.66, 245.64) (228.66, 245.64)      /C0_2 を	<|special_separator|>
 (238.66, 235.01) (248.66, 235.01) (248.66, 245.64) (238.66, 245.64)      /C0_2 押	<|special_separator|>
 (248.66, 235.01) (258.66, 235.01) (258.66, 245.64) (248.66, 245.64)      /C0_2 す	<|special_separator|>
@@ -642,7 +642,7 @@
 (285.84, 219.16) (291.51, 219.16) (291.51, 228.88) (285.84, 228.88)      /C0_1 3	<|special_separator|>
 (291.50, 219.16) (297.17, 219.16) (297.17, 228.88) (291.50, 228.88)      /C0_1 3	<|special_separator|>
 (297.16, 219.16) (299.99, 219.16) (299.99, 228.88) (297.16, 228.88)      /C0_1 ,	<|special_separator|>
-(299.97, 219.16) (302.97, 219.16) (302.97, 228.88) (299.97, 228.88)      /C0_1 		<|special_separator|>
+(299.97, 219.16) (302.97, 219.16) (302.97, 228.88) (299.97, 228.88)      /C0_1  	<|special_separator|>
 (302.97, 219.16) (308.64, 219.16) (308.64, 228.88) (302.97, 228.88)      /C0_1 3	<|special_separator|>
 (308.62, 219.16) (314.29, 219.16) (314.29, 228.88) (308.62, 228.88)      /C0_1 4	<|special_separator|>
 (314.28, 219.66) (321.28, 219.66) (321.28, 227.22) (314.28, 227.22)      /C0_1 ペ	<|special_separator|>
@@ -733,13 +733,13 @@
 (478.51, 692.42) (488.51, 692.42) (488.51, 703.05) (478.51, 703.05)      /C0_2 ん	<|special_separator|>
 (488.51, 692.42) (498.51, 692.42) (498.51, 703.05) (488.51, 703.05)      /C0_2 。	<|special_separator|>
 (493.51, 692.42) (503.51, 692.42) (503.51, 703.05) (493.51, 703.05)      /C0_2 ）	<|special_separator|>
-(265.01, 671.57) (270.01, 671.57) (270.01, 687.76) (265.01, 687.76)      /C0_1 		<|special_separator|>
+(265.01, 671.57) (270.01, 671.57) (270.01, 687.76) (265.01, 687.76)      /C0_1  	<|special_separator|>
 (270.01, 671.57) (285.01, 671.57) (285.01, 687.76) (270.01, 687.76)      /C0_1 を	<|special_separator|>
 (285.01, 671.57) (300.01, 671.57) (300.01, 687.76) (285.01, 687.76)      /C0_1 押	<|special_separator|>
 (300.01, 671.57) (315.01, 671.57) (315.01, 687.76) (300.01, 687.76)      /C0_1 す	<|special_separator|>
 (315.01, 671.57) (330.01, 671.57) (330.01, 687.76) (315.01, 687.76)      /C0_1 。	<|special_separator|>
 (185.00, 654.19) (192.00, 654.19) (192.00, 661.64) (185.00, 661.64)      /C0_2 ●	<|special_separator|>
-(192.00, 652.35) (195.33, 652.35) (195.33, 662.99) (192.00, 662.99)      /C0_2 	<|special_separator|>
+(192.00, 652.35) (195.33, 652.35) (195.33, 662.99) (192.00, 662.99)      /C0_2 		<|special_separator|>
 (194.85, 652.35) (204.85, 652.35) (204.85, 662.99) (194.85, 662.99)      /C0_2 室	<|special_separator|>
 (204.64, 652.35) (214.64, 652.35) (214.64, 662.99) (204.64, 662.99)      /C0_2 内	<|special_separator|>
 (214.15, 652.35) (224.15, 652.35) (224.15, 662.99) (214.15, 662.99)      /C0_2 ・	<|special_separator|>
@@ -870,7 +870,7 @@
 (527.21, 603.46) (535.76, 603.46) (535.76, 613.04) (527.21, 613.04)      /C0_2 場	<|special_separator|>
 (535.60, 603.46) (544.15, 603.46) (544.15, 613.04) (535.60, 613.04)      /C0_2 合	<|special_separator|>
 (185.00, 578.52) (192.00, 578.52) (192.00, 585.97) (185.00, 585.97)      /C0_2 ●	<|special_separator|>
-(192.00, 576.68) (195.33, 576.68) (195.33, 587.32) (192.00, 587.32)      /C0_2 	<|special_separator|>
+(192.00, 576.68) (195.33, 576.68) (195.33, 587.32) (192.00, 587.32)      /C0_2 		<|special_separator|>
 (194.92, 576.68) (204.72, 576.68) (204.72, 587.32) (194.92, 587.32)      /C0_2 人	<|special_separator|>
 (204.72, 576.68) (214.52, 576.68) (214.52, 587.32) (204.72, 587.32)      /C0_2 ・	<|special_separator|>
 (214.52, 576.68) (224.32, 576.68) (224.32, 587.32) (214.52, 587.32)      /C0_2 床	<|special_separator|>
@@ -932,7 +932,7 @@
 (488.09, 564.85) (493.76, 564.85) (493.76, 574.57) (488.09, 574.57)      /C0_1 3	<|special_separator|>
 (493.75, 564.85) (499.42, 564.85) (499.42, 574.57) (493.75, 574.57)      /C0_1 5	<|special_separator|>
 (499.40, 564.85) (502.23, 564.85) (502.23, 574.57) (499.40, 574.57)      /C0_1 ,	<|special_separator|>
-(502.21, 564.85) (505.21, 564.85) (505.21, 574.57) (502.21, 574.57)      /C0_1 		<|special_separator|>
+(502.21, 564.85) (505.21, 564.85) (505.21, 574.57) (502.21, 574.57)      /C0_1  	<|special_separator|>
 (505.20, 564.85) (510.87, 564.85) (510.87, 574.57) (505.20, 574.57)      /C0_1 4	<|special_separator|>
 (510.85, 564.85) (516.52, 564.85) (516.52, 574.57) (510.85, 574.57)      /C0_1 0	<|special_separator|>
 (516.52, 565.35) (523.52, 565.35) (523.52, 572.91) (516.52, 572.91)      /C0_1 ペ	<|special_separator|>

groundtruth/parser/cropbox_versus_mediabox_01.pdf.page_no_1.py.json.line.txt

@@ -1,9 +1,9 @@
 (559.67, 010.43) (582.69, 010.43) (582.69, 030.93) (559.67, 030.93)       /TT0 11	<|special_separator|>
 (034.02, 772.75) (466.02, 772.75) (466.02, 809.38) (034.02, 809.38)      /C0_0 エアコンに運転をまかせる	<|special_separator|>
 (038.27, 166.09) (078.27, 166.09) (078.27, 176.89) (038.27, 176.89)      /C0_1 お知らせ	<|special_separator|>
-(034.02, 152.53) (160.09, 150.87) (160.09, 160.45) (034.02, 159.23)      /C0_2 ●ご購入時は自動内部クリーン	<|special_separator|>
+(034.02, 152.53) (160.09, 150.87) (160.09, 160.45) (034.02, 159.23)      /C0_2 ●	ご購入時は自動内部クリーン	<|special_separator|>
 (038.59, 138.37) (088.99, 138.37) (088.99, 147.95) (038.59, 147.95)      /C0_2 「入」 です。	<|special_separator|>
-(034.02, 127.53) (168.97, 125.87) (168.97, 135.45) (034.02, 134.23)      /C0_2 ●自動内部クリーン 「入」 の場合、	<|special_separator|>
+(034.02, 127.53) (168.97, 125.87) (168.97, 135.45) (034.02, 134.23)      /C0_2 ●	自動内部クリーン 「入」 の場合、	<|special_separator|>
 (043.08, 112.75) (052.81, 112.75) (052.81, 122.13) (043.08, 122.13)       /TT1 AI	<|special_separator|>
 (052.81, 113.37) (160.81, 113.37) (160.81, 122.95) (052.81, 122.95)      /C0_2 快適自動運転で選択された	<|special_separator|>
 (043.05, 100.87) (163.20, 100.87) (163.20, 110.44) (043.05, 110.44)      /C0_2 冷房・除湿・除湿冷房運転を停	<|special_separator|>
@@ -12,8 +12,8 @@
 (121.11, 075.75) (159.74, 075.43) (159.74, 082.99) (121.11, 084.06)      /C0_1 ▶28ページ	<|special_separator|>
 (164.41, 075.19) (167.41, 075.19) (167.41, 084.77) (164.41, 084.77)      /C0_2 	<|special_separator|>
 (042.74, 062.02) (163.21, 062.02) (163.21, 071.59) (042.74, 071.59)      /C0_2 自動内部クリーンを停止したい	<|special_separator|>
-(043.03, 047.50) (115.53, 047.50) (115.53, 057.08) (043.03, 057.08)      /C0_2 場合は、もう一度	<|special_separator|>
-(146.47, 047.50) (158.29, 047.50) (158.29, 057.07) (146.47, 057.07)      /C0_2 を	<|special_separator|>
+(043.03, 047.50) (115.53, 047.50) (115.53, 057.08) (043.03, 057.08)      /C0_2 場合は、もう一度 	<|special_separator|>
+(146.47, 047.50) (158.29, 047.50) (158.29, 057.07) (146.47, 057.07)      /C0_2  を	<|special_separator|>
 (043.02, 032.98) (111.90, 032.98) (111.90, 042.56) (043.02, 042.56)      /C0_2 押してください。	<|special_separator|>
 (180.00, 049.56) (210.00, 049.56) (210.00, 060.36) (180.00, 060.36)      /C0_1 お願い	<|special_separator|>
 (175.75, 033.66) (400.75, 033.66) (400.75, 043.23) (175.75, 043.23)      /C0_2 お好みに合わないときは運転モードを変えてください。	<|special_separator|>
@@ -21,36 +21,36 @@
 (065.44, 192.86) (141.94, 192.86) (141.94, 202.44) (065.44, 202.44)      /C0_3 運転ランプ （緑色）	<|special_separator|>
 (188.59, 174.18) (202.33, 174.18) (202.33, 185.62) (188.59, 185.62)       /TT2 AI	<|special_separator|>
 (202.33, 174.79) (290.33, 174.79) (290.33, 186.62) (202.33, 186.62)      /C0_4 快適自動運転とは	<|special_separator|>
-(188.59, 163.26) (446.84, 161.41) (446.84, 172.05) (188.59, 170.70)      /C0_2 ●壁温度と室内温度に応じて、室内温度制御を行います。 	<|special_separator|>
+(188.59, 163.26) (446.84, 161.41) (446.84, 172.05) (188.59, 170.70)      /C0_2 ●	壁温度と室内温度に応じて、室内温度制御を行います。 	<|special_separator|>
 (198.45, 149.41) (542.01, 149.41) (542.01, 160.05) (198.45, 160.05)      /C0_2 壁温度が高い場合、冷房時は低めの室内温度に、壁温度が低い場合、暖房時は	<|special_separator|>
 (198.51, 137.41) (388.51, 137.41) (388.51, 148.05) (198.51, 148.05)      /C0_2 高めの室内温度になるように運転します。	<|special_separator|>
-(188.59, 124.42) (326.84, 122.58) (326.84, 133.22) (188.59, 131.87)      /C0_2 ●お好みの運転を学習します。 	<|special_separator|>
+(188.59, 124.42) (326.84, 122.58) (326.84, 133.22) (188.59, 131.87)      /C0_2 ●	お好みの運転を学習します。 	<|special_separator|>
 (198.46, 110.58) (542.51, 110.58) (542.51, 121.22) (198.46, 121.22)      /C0_2 設定温度と室内温度、壁温度から快適と感じるお好みの設定を蓄積し、最適な	<|special_separator|>
 (198.51, 098.58) (356.84, 098.58) (356.84, 109.22) (198.51, 109.22)      /C0_2 運転モードを予測して運転します。 	<|special_separator|>
 (193.42, 086.58) (542.48, 086.58) (542.48, 097.22) (193.42, 097.22)      /C0_2 （ 「快適自動運転」 の場合は、床温度、壁温度の検知や、お好みの設定の蓄積は	<|special_separator|>
 (198.50, 074.58) (263.50, 074.58) (263.50, 085.22) (198.50, 085.22)      /C0_2 行いません。 ）	<|special_separator|>
 (185.00, 436.94) (297.00, 436.94) (297.00, 453.61) (185.00, 453.61)      /C0_5 停止したいとき	<|special_separator|>
-(226.27, 411.01) (291.26, 411.01) (291.26, 427.21) (226.27, 427.21)      /C0_1 	を押す。	<|special_separator|>
-(185.00, 393.97) (295.92, 392.13) (295.92, 402.77) (185.00, 401.42)      /C0_2 ● 「停止」 と表示した後、	<|special_separator|>
+(226.27, 411.01) (291.26, 411.01) (291.26, 427.21) (226.27, 427.21)      /C0_1  を押す。	<|special_separator|>
+(185.00, 393.97) (295.92, 392.13) (295.92, 402.77) (185.00, 401.42)      /C0_2 ●	 「停止」 と表示した後、	<|special_separator|>
 (194.92, 380.13) (294.92, 380.13) (294.92, 390.77) (194.92, 390.77)      /C0_2 リモコン表示が消灯し	<|special_separator|>
 (194.92, 368.13) (224.92, 368.13) (224.92, 378.77) (194.92, 378.77)      /C0_2 ます。	<|special_separator|>
-(185.00, 354.30) (374.92, 352.46) (374.92, 363.10) (185.00, 361.75)      /C0_2 ●自動内部クリーンを停止したい場合は、	<|special_separator|>
-(194.92, 338.94) (238.25, 338.94) (238.25, 349.58) (194.92, 349.58)      /C0_2 もう一度	<|special_separator|>
-(269.20, 338.94) (362.53, 338.94) (362.53, 349.58) (269.20, 349.58)      /C0_2 を押してください。	<|special_separator|>
+(185.00, 354.30) (374.92, 352.46) (374.92, 363.10) (185.00, 361.75)      /C0_2 ●	自動内部クリーンを停止したい場合は、	<|special_separator|>
+(194.92, 338.94) (238.25, 338.94) (238.25, 349.58) (194.92, 349.58)      /C0_2 もう一度 	<|special_separator|>
+(269.20, 338.94) (362.53, 338.94) (362.53, 349.58) (269.20, 349.58)      /C0_2  を押してください。	<|special_separator|>
 (397.60, 436.94) (541.60, 436.94) (541.60, 453.61) (397.60, 453.61)      /C0_5 風向を変えたいとき	<|special_separator|>
 (402.28, 419.00) (461.26, 418.68) (461.26, 426.24) (402.28, 427.32)      /C0_1 ▶16～19ページ	<|special_separator|>
 (185.00, 522.47) (361.00, 522.47) (361.00, 539.14) (185.00, 539.14)      /C0_5 設定温度を変えたいとき	<|special_separator|>
-(208.54, 486.42) (273.54, 486.42) (273.54, 502.62) (208.54, 502.62)      /C0_1 	を押す。	<|special_separator|>
-(284.21, 506.09) (543.47, 504.25) (543.47, 514.89) (284.21, 513.54)      /C0_2 ●適温 ※ から、ー5.0℃~＋5.0℃の範囲で変更できます。 	<|special_separator|>
+(208.54, 486.42) (273.54, 486.42) (273.54, 502.62) (208.54, 502.62)      /C0_1  を押す。	<|special_separator|>
+(284.21, 506.09) (543.47, 504.25) (543.47, 514.89) (284.21, 513.54)      /C0_2 ●	適温 ※ から、ー5.0℃ ~ ＋5.0℃の範囲で変更できます。 	<|special_separator|>
 (289.14, 492.25) (362.65, 492.25) (362.65, 502.89) (289.14, 502.89)      /C0_2 （0.5℃きざみ） 	<|special_separator|>
-(293.76, 480.25) (549.57, 480.25) (549.57, 490.89) (293.76, 490.89)      /C0_2 ※ 適温とは、 室内・屋外温度よりエアコンが自動で決定した温度です。	<|special_separator|>
+(293.76, 480.25) (549.57, 480.25) (549.57, 490.89) (293.76, 490.89)      /C0_2 ※ 適温とは、 室内・屋外温度よりエアコンが自動で決定した温度です。	<|special_separator|>
 (185.00, 301.98) (377.00, 301.98) (377.00, 318.65) (185.00, 318.65)      /C0_5 運転状態を確認したいとき	<|special_separator|>
-(185.00, 284.90) (343.25, 283.05) (343.25, 293.69) (185.00, 292.34)      /C0_2 ●エアコンが選択した運転モード、 	<|special_separator|>
+(185.00, 284.90) (343.25, 283.05) (343.25, 293.69) (185.00, 292.34)      /C0_2 ●	エアコンが選択した運転モード、 	<|special_separator|>
 (194.92, 271.05) (253.25, 271.05) (253.25, 281.69) (194.92, 281.69)      /C0_2 設定温度は、 	<|special_separator|>
-(227.41, 255.04) (306.99, 255.04) (306.99, 265.68) (227.41, 265.68)      /C0_2  で確認できます。 	<|special_separator|>
-(227.41, 235.01) (321.99, 235.01) (321.99, 245.64) (227.41, 245.64)      /C0_2  を押すごとに表示が	<|special_separator|>
+(227.41, 255.04) (306.99, 255.04) (306.99, 265.68) (227.41, 265.68)      /C0_2   で確認できます。 	<|special_separator|>
+(227.41, 235.01) (321.99, 235.01) (321.99, 245.64) (227.41, 245.64)      /C0_2   を押すごとに表示が	<|special_separator|>
 (194.92, 218.82) (274.92, 218.82) (274.92, 229.45) (194.92, 229.45)      /C0_2 切り換わります。	<|special_separator|>
-(279.57, 219.98) (335.24, 219.66) (335.24, 227.22) (279.57, 228.30)      /C0_1 ▶33,	34ページ	<|special_separator|>
+(279.57, 219.98) (335.24, 219.66) (335.24, 227.22) (279.57, 228.30)      /C0_1 ▶33, 34ページ	<|special_separator|>
 (509.54, 208.77) (533.54, 208.77) (533.54, 215.15) (509.54, 215.15)      /C0_6 室内湿度	<|special_separator|>
 (453.17, 208.77) (477.17, 208.77) (477.17, 215.15) (453.17, 215.15)      /C0_6 室内温度	<|special_separator|>
 (396.38, 208.77) (420.38, 208.77) (420.38, 215.15) (396.38, 215.15)      /C0_6 設定温度	<|special_separator|>
@@ -58,8 +58,8 @@
 (214.98, 719.44) (358.98, 719.44) (358.98, 744.44) (214.98, 744.44)      /C0_5 快適自動運転	<|special_separator|>
 (193.51, 704.42) (523.51, 704.42) (523.51, 715.05) (193.51, 715.05)      /C0_2 お好みの設定を学習し、温度、湿度、気流をエアコンが自動で選択して、	<|special_separator|>
 (193.51, 692.42) (503.51, 692.42) (503.51, 703.05) (193.51, 703.05)      /C0_2 快適な運転を行います。 （暖房時の湿度コントロールはありません。 ）	<|special_separator|>
-(265.01, 671.57) (330.01, 671.57) (330.01, 687.76) (265.01, 687.76)      /C0_1 	を押す。	<|special_separator|>
-(185.00, 654.19) (273.26, 652.35) (273.26, 662.99) (185.00, 661.64)      /C0_2 ●室内・屋外温度や	<|special_separator|>
+(265.01, 671.57) (330.01, 671.57) (330.01, 687.76) (265.01, 687.76)      /C0_1  を押す。	<|special_separator|>
+(185.00, 654.19) (273.26, 652.35) (273.26, 662.99) (185.00, 661.64)      /C0_2 ●	室内・屋外温度や	<|special_separator|>
 (273.13, 651.66) (283.80, 651.66) (283.80, 662.08) (273.13, 662.08)       /TT1 AI	<|special_separator|>
 (283.63, 652.35) (392.00, 652.35) (392.00, 662.99) (283.63, 662.99)      /C0_2 快適自動運転変更前の設	<|special_separator|>
 (194.86, 640.35) (397.21, 640.35) (397.21, 650.99) (194.86, 650.99)      /C0_2 定内容に応じて、自動で運転モード （暖房、	<|special_separator|>
@@ -73,11 +73,11 @@
 (399.13, 592.47) (474.16, 592.46) (474.16, 602.04) (399.13, 602.04)      /C0_2 （ご購入時の設定）	<|special_separator|>
 (489.75, 614.46) (532.50, 614.46) (532.50, 624.04) (489.75, 624.04)      /C0_2 学習制御を	<|special_separator|>
 (478.11, 603.46) (544.15, 603.46) (544.15, 613.04) (478.11, 613.04)      /C0_2 行っていない場合	<|special_separator|>
-(185.00, 578.52) (334.08, 576.68) (334.08, 587.32) (185.00, 585.97)      /C0_2 ●人・床センサー 「固定」 の場合、	<|special_separator|>
+(185.00, 578.52) (334.08, 576.68) (334.08, 587.32) (185.00, 585.97)      /C0_2 ●	人・床センサー 「固定」 の場合、	<|special_separator|>
 (334.08, 575.99) (344.89, 575.99) (344.89, 586.41) (334.08, 586.41)       /TT1 AI	<|special_separator|>
 (344.89, 576.68) (529.46, 576.68) (529.46, 587.32) (344.89, 587.32)      /C0_2 快適自動運転時には、学習制御を行わず、 	<|special_separator|>
 (194.92, 562.71) (376.61, 562.71) (376.61, 573.35) (194.92, 573.35)      /C0_2 リモコンに 「快適自動」 と表示されます。	<|special_separator|>
-(481.82, 565.67) (537.49, 565.35) (537.49, 572.91) (481.82, 573.98)      /C0_1 ▶35,	40ページ	<|special_separator|>
+(481.82, 565.67) (537.49, 565.35) (537.49, 572.91) (481.82, 573.98)      /C0_1 ▶35, 40ページ	<|special_separator|>
 (578.96, 438.06) (589.96, 438.06) (589.96, 449.26) (578.96, 449.26)      /C0_0 便	<|special_separator|>
 (578.96, 278.74) (589.96, 278.74) (589.96, 289.93) (578.96, 289.93)      /C0_0 お	<|special_separator|>
 (578.96, 153.54) (589.96, 153.54) (589.96, 164.74) (578.96, 164.74)      /C0_0 ࠔ	<|special_separator|>

groundtruth/parser/cropbox_versus_mediabox_01.pdf.page_no_1.py.json.word.txt

@@ -1,10 +1,12 @@
 (559.67, 010.43) (582.69, 010.43) (582.69, 030.93) (559.67, 030.93)       /TT0 11	<|special_separator|>
 (034.02, 772.75) (466.02, 772.75) (466.02, 809.38) (034.02, 809.38)      /C0_0 エアコンに運転をまかせる	<|special_separator|>
 (038.27, 166.09) (078.27, 166.09) (078.27, 176.89) (038.27, 176.89)      /C0_1 お知らせ	<|special_separator|>
-(034.02, 152.53) (160.09, 150.87) (160.09, 160.45) (034.02, 159.23)      /C0_2 ●ご購入時は自動内部クリーン	<|special_separator|>
+(034.02, 152.53) (040.32, 152.53) (040.32, 159.23) (034.02, 159.23)      /C0_2 ●	<|special_separator|>
+(043.09, 150.87) (160.09, 150.87) (160.09, 160.45) (043.09, 160.45)      /C0_2 ご購入時は自動内部クリーン	<|special_separator|>
 (038.59, 138.37) (065.59, 138.37) (065.59, 147.95) (038.59, 147.95)      /C0_2 「入」	<|special_separator|>
 (061.99, 138.37) (088.99, 138.37) (088.99, 147.95) (061.99, 147.95)      /C0_2 です。	<|special_separator|>
-(034.02, 127.53) (114.45, 125.87) (114.45, 135.45) (034.02, 134.23)      /C0_2 ●自動内部クリーン	<|special_separator|>
+(034.02, 127.53) (040.32, 127.53) (040.32, 134.23) (034.02, 134.23)      /C0_2 ●	<|special_separator|>
+(043.02, 125.87) (114.45, 125.87) (114.45, 135.45) (043.02, 135.45)      /C0_2 自動内部クリーン	<|special_separator|>
 (109.85, 125.87) (136.78, 125.87) (136.78, 135.45) (109.85, 135.45)      /C0_2 「入」	<|special_separator|>
 (133.08, 125.87) (168.97, 125.87) (168.97, 135.45) (133.08, 135.45)      /C0_2 の場合、	<|special_separator|>
 (043.08, 112.75) (052.81, 112.75) (052.81, 122.13) (043.08, 122.13)       /TT1 AI	<|special_separator|>
@@ -15,8 +17,8 @@
 (121.11, 075.75) (159.74, 075.43) (159.74, 082.99) (121.11, 084.06)      /C0_1 ▶28ページ	<|special_separator|>
 (164.41, 075.19) (167.41, 075.19) (167.41, 084.77) (164.41, 084.77)      /C0_2 	<|special_separator|>
 (042.74, 062.02) (163.21, 062.02) (163.21, 071.59) (042.74, 071.59)      /C0_2 自動内部クリーンを停止したい	<|special_separator|>
-(043.03, 047.50) (115.53, 047.50) (115.53, 057.08) (043.03, 057.08)      /C0_2 場合は、もう一度	<|special_separator|>
-(146.47, 047.50) (158.29, 047.50) (158.29, 057.07) (146.47, 057.07)      /C0_2 を	<|special_separator|>
+(043.03, 047.50) (112.64, 047.50) (112.64, 057.08) (043.03, 057.08)      /C0_2 場合は、もう一度	<|special_separator|>
+(149.29, 047.50) (158.29, 047.50) (158.29, 057.07) (149.29, 057.07)      /C0_2 を	<|special_separator|>
 (043.02, 032.98) (111.90, 032.98) (111.90, 042.56) (043.02, 042.56)      /C0_2 押してください。	<|special_separator|>
 (180.00, 049.56) (210.00, 049.56) (210.00, 060.36) (180.00, 060.36)      /C0_1 お願い	<|special_separator|>
 (175.75, 033.66) (400.75, 033.66) (400.75, 043.23) (175.75, 043.23)      /C0_2 お好みに合わないときは運転モードを変えてください。	<|special_separator|>
@@ -25,11 +27,13 @@
 (105.94, 192.86) (141.94, 192.86) (141.94, 202.44) (105.94, 202.44)      /C0_3 （緑色）	<|special_separator|>
 (188.59, 174.18) (202.33, 174.18) (202.33, 185.62) (188.59, 185.62)       /TT2 AI	<|special_separator|>
 (202.33, 174.79) (290.33, 174.79) (290.33, 186.62) (202.33, 186.62)      /C0_4 快適自動運転とは	<|special_separator|>
-(188.59, 163.26) (448.51, 161.41) (448.51, 172.05) (188.59, 170.70)      /C0_2 ●壁温度と室内温度に応じて、室内温度制御を行います。	<|special_separator|>
+(188.59, 163.26) (195.59, 163.26) (195.59, 170.70) (188.59, 170.70)      /C0_2 ●	<|special_separator|>
+(198.51, 161.41) (448.51, 161.41) (448.51, 172.05) (198.51, 172.05)      /C0_2 壁温度と室内温度に応じて、室内温度制御を行います。	<|special_separator|>
 (443.51, 161.41) (446.84, 161.41) (446.84, 172.05) (443.51, 172.05)      /C0_2 	<|special_separator|>
 (198.45, 149.41) (542.01, 149.41) (542.01, 160.05) (198.45, 160.05)      /C0_2 壁温度が高い場合、冷房時は低めの室内温度に、壁温度が低い場合、暖房時は	<|special_separator|>
 (198.51, 137.41) (388.51, 137.41) (388.51, 148.05) (198.51, 148.05)      /C0_2 高めの室内温度になるように運転します。	<|special_separator|>
-(188.59, 124.42) (328.51, 122.58) (328.51, 133.22) (188.59, 131.87)      /C0_2 ●お好みの運転を学習します。	<|special_separator|>
+(188.59, 124.42) (195.59, 124.42) (195.59, 131.87) (188.59, 131.87)      /C0_2 ●	<|special_separator|>
+(198.51, 122.58) (328.51, 122.58) (328.51, 133.22) (198.51, 133.22)      /C0_2 お好みの運転を学習します。	<|special_separator|>
 (323.51, 122.58) (326.84, 122.58) (326.84, 133.22) (323.51, 133.22)      /C0_2 	<|special_separator|>
 (198.46, 110.58) (542.51, 110.58) (542.51, 121.22) (198.46, 121.22)      /C0_2 設定温度と室内温度、壁温度から快適と感じるお好みの設定を蓄積し、最適な	<|special_separator|>
 (198.51, 098.58) (358.51, 098.58) (358.51, 109.22) (198.51, 109.22)      /C0_2 運転モードを予測して運転します。	<|special_separator|>
@@ -41,35 +45,40 @@
 (253.50, 074.58) (263.50, 074.58) (263.50, 085.22) (253.50, 085.22)      /C0_2 ）	<|special_separator|>
 (185.00, 436.94) (297.00, 436.94) (297.00, 453.61) (185.00, 453.61)      /C0_5 停止したいとき	<|special_separator|>
 (231.26, 411.01) (291.26, 411.01) (291.26, 427.21) (231.26, 427.21)      /C0_1 を押す。	<|special_separator|>
-(185.00, 393.97) (195.33, 392.13) (195.33, 402.77) (185.00, 401.42)      /C0_2 ●	<|special_separator|>
+(185.00, 393.97) (192.00, 393.97) (192.00, 401.42) (185.00, 401.42)      /C0_2 ●	<|special_separator|>
 (189.92, 392.13) (229.92, 392.13) (229.92, 402.77) (189.92, 402.77)      /C0_2 「停止」	<|special_separator|>
 (225.92, 392.13) (295.92, 392.13) (295.92, 402.77) (225.92, 402.77)      /C0_2 と表示した後、	<|special_separator|>
 (194.92, 380.13) (294.92, 380.13) (294.92, 390.77) (194.92, 390.77)      /C0_2 リモコン表示が消灯し	<|special_separator|>
 (194.92, 368.13) (224.92, 368.13) (224.92, 378.77) (194.92, 378.77)      /C0_2 ます。	<|special_separator|>
-(185.00, 354.30) (374.92, 352.46) (374.92, 363.10) (185.00, 361.75)      /C0_2 ●自動内部クリーンを停止したい場合は、	<|special_separator|>
-(194.92, 338.94) (238.25, 338.94) (238.25, 349.58) (194.92, 349.58)      /C0_2 もう一度	<|special_separator|>
-(269.20, 338.94) (362.53, 338.94) (362.53, 349.58) (269.20, 349.58)      /C0_2 を押してください。	<|special_separator|>
+(185.00, 354.30) (192.00, 354.30) (192.00, 361.75) (185.00, 361.75)      /C0_2 ●	<|special_separator|>
+(194.92, 352.46) (374.92, 352.46) (374.92, 363.10) (194.92, 363.10)      /C0_2 自動内部クリーンを停止したい場合は、	<|special_separator|>
+(194.92, 338.94) (234.92, 338.94) (234.92, 349.58) (194.92, 349.58)      /C0_2 もう一度	<|special_separator|>
+(272.53, 338.94) (362.53, 338.94) (362.53, 349.58) (272.53, 349.58)      /C0_2 を押してください。	<|special_separator|>
 (397.60, 436.94) (541.60, 436.94) (541.60, 453.61) (397.60, 453.61)      /C0_5 風向を変えたいとき	<|special_separator|>
 (402.28, 419.00) (461.26, 418.68) (461.26, 426.24) (402.28, 427.32)      /C0_1 ▶16～19ページ	<|special_separator|>
 (185.00, 522.47) (361.00, 522.47) (361.00, 539.14) (185.00, 539.14)      /C0_5 設定温度を変えたいとき	<|special_separator|>
 (213.54, 486.42) (273.54, 486.42) (273.54, 502.62) (213.54, 502.62)      /C0_1 を押す。	<|special_separator|>
-(284.21, 506.09) (314.13, 504.25) (314.13, 514.89) (284.21, 513.54)      /C0_2 ●適温	<|special_separator|>
+(284.21, 506.09) (291.21, 506.09) (291.21, 513.54) (284.21, 513.54)      /C0_2 ●	<|special_separator|>
+(294.13, 504.25) (314.13, 504.25) (314.13, 514.89) (294.13, 514.89)      /C0_2 適温	<|special_separator|>
 (314.14, 510.14) (319.97, 510.14) (319.97, 516.34) (314.14, 516.34)      /C0_2 ※	<|special_separator|>
-(319.97, 504.25) (545.14, 504.25) (545.14, 514.89) (319.97, 514.89)      /C0_2 から、ー5.0℃~＋5.0℃の範囲で変更できます。	<|special_separator|>
+(319.97, 504.25) (385.15, 504.25) (385.15, 514.89) (319.97, 514.89)      /C0_2 から、ー5.0℃	<|special_separator|>
+(388.48, 504.25) (396.63, 504.25) (396.63, 514.89) (388.48, 514.89)      /C0_2 ~	<|special_separator|>
+(399.96, 504.25) (545.14, 504.25) (545.14, 514.89) (399.96, 514.89)      /C0_2 ＋5.0℃の範囲で変更できます。	<|special_separator|>
 (540.14, 504.25) (543.47, 504.25) (543.47, 514.89) (540.14, 514.89)      /C0_2 	<|special_separator|>
 (289.14, 492.25) (364.32, 492.25) (364.32, 502.89) (289.14, 502.89)      /C0_2 （0.5℃きざみ）	<|special_separator|>
 (359.32, 492.25) (362.65, 492.25) (362.65, 502.89) (359.32, 502.89)      /C0_2 	<|special_separator|>
-(293.76, 480.25) (304.77, 480.25) (304.77, 490.89) (293.76, 490.89)      /C0_2 ※	<|special_separator|>
+(293.76, 480.25) (304.77, 480.25) (304.77, 490.89) (293.76, 490.89)      /C0_2 ※	<|special_separator|>
 (301.82, 480.25) (343.13, 480.25) (343.13, 490.89) (301.82, 490.89)      /C0_2 適温とは、	<|special_separator|>
 (339.16, 480.25) (549.57, 480.25) (549.57, 490.89) (339.16, 490.89)      /C0_2 室内・屋外温度よりエアコンが自動で決定した温度です。	<|special_separator|>
 (185.00, 301.98) (377.00, 301.98) (377.00, 318.65) (185.00, 318.65)      /C0_5 運転状態を確認したいとき	<|special_separator|>
-(185.00, 284.90) (344.92, 283.05) (344.92, 293.69) (185.00, 292.34)      /C0_2 ●エアコンが選択した運転モード、	<|special_separator|>
+(185.00, 284.90) (192.00, 284.90) (192.00, 292.34) (185.00, 292.34)      /C0_2 ●	<|special_separator|>
+(194.92, 283.05) (344.92, 283.05) (344.92, 293.69) (194.92, 293.69)      /C0_2 エアコンが選択した運転モード、	<|special_separator|>
 (339.92, 283.05) (343.25, 283.05) (343.25, 293.69) (339.92, 293.69)      /C0_2 	<|special_separator|>
 (194.92, 271.05) (254.92, 271.05) (254.92, 281.69) (194.92, 281.69)      /C0_2 設定温度は、	<|special_separator|>
 (249.92, 271.05) (253.25, 271.05) (253.25, 281.69) (249.92, 281.69)      /C0_2 	<|special_separator|>
-(227.41, 255.04) (308.66, 255.04) (308.66, 265.68) (227.41, 265.68)      /C0_2 で確認できます。	<|special_separator|>
+(228.66, 255.04) (308.66, 255.04) (308.66, 265.68) (228.66, 265.68)      /C0_2 で確認できます。	<|special_separator|>
 (303.66, 255.04) (306.99, 255.04) (306.99, 265.68) (303.66, 265.68)      /C0_2 	<|special_separator|>
-(227.41, 235.01) (321.99, 235.01) (321.99, 245.64) (227.41, 245.64)      /C0_2 を押すごとに表示が	<|special_separator|>
+(228.66, 235.01) (321.99, 235.01) (321.99, 245.64) (228.66, 245.64)      /C0_2 を押すごとに表示が	<|special_separator|>
 (194.92, 218.82) (274.92, 218.82) (274.92, 229.45) (194.92, 229.45)      /C0_2 切り換わります。	<|special_separator|>
 (279.57, 219.98) (299.99, 219.16) (299.99, 228.88) (279.57, 228.30)      /C0_1 ▶33,	<|special_separator|>
 (302.97, 219.16) (335.24, 219.66) (335.24, 227.22) (302.97, 228.88)      /C0_1 34ページ	<|special_separator|>
@@ -83,7 +92,8 @@
 (298.51, 692.42) (498.51, 692.42) (498.51, 703.05) (298.51, 703.05)      /C0_2 （暖房時の湿度コントロールはありません。	<|special_separator|>
 (493.51, 692.42) (503.51, 692.42) (503.51, 703.05) (493.51, 703.05)      /C0_2 ）	<|special_separator|>
 (270.01, 671.57) (330.01, 671.57) (330.01, 687.76) (270.01, 687.76)      /C0_1 を押す。	<|special_separator|>
-(185.00, 654.19) (273.26, 652.35) (273.26, 662.99) (185.00, 661.64)      /C0_2 ●室内・屋外温度や	<|special_separator|>
+(185.00, 654.19) (192.00, 654.19) (192.00, 661.64) (185.00, 661.64)      /C0_2 ●	<|special_separator|>
+(194.85, 652.35) (273.26, 652.35) (273.26, 662.99) (194.85, 662.99)      /C0_2 室内・屋外温度や	<|special_separator|>
 (273.13, 651.66) (283.80, 651.66) (283.80, 662.08) (273.13, 662.08)       /TT1 AI	<|special_separator|>
 (283.63, 652.35) (392.00, 652.35) (392.00, 662.99) (283.63, 662.99)      /C0_2 快適自動運転変更前の設	<|special_separator|>
 (194.86, 640.35) (360.81, 640.35) (360.81, 650.99) (194.86, 650.99)      /C0_2 定内容に応じて、自動で運転モード	<|special_separator|>
@@ -99,7 +109,8 @@
 (399.13, 592.47) (474.16, 592.46) (474.16, 602.04) (399.13, 602.04)      /C0_2 （ご購入時の設定）	<|special_separator|>
 (489.75, 614.46) (532.50, 614.46) (532.50, 624.04) (489.75, 624.04)      /C0_2 学習制御を	<|special_separator|>
 (478.11, 603.46) (544.15, 603.46) (544.15, 613.04) (478.11, 613.04)      /C0_2 行っていない場合	<|special_separator|>
-(185.00, 578.52) (263.52, 576.68) (263.52, 587.32) (185.00, 585.97)      /C0_2 ●人・床センサー	<|special_separator|>
+(185.00, 578.52) (192.00, 578.52) (192.00, 585.97) (185.00, 585.97)      /C0_2 ●	<|special_separator|>
+(194.92, 576.68) (263.52, 576.68) (263.52, 587.32) (194.92, 587.32)      /C0_2 人・床センサー	<|special_separator|>
 (259.60, 576.68) (298.80, 576.68) (298.80, 587.32) (259.60, 587.32)      /C0_2 「固定」	<|special_separator|>
 (294.88, 576.68) (334.08, 576.68) (334.08, 587.32) (294.88, 587.32)      /C0_2 の場合、	<|special_separator|>
 (334.08, 575.99) (344.89, 575.99) (344.89, 586.41) (334.08, 586.41)       /TT1 AI	<|special_separator|>

groundtruth/parser/fillable_form.pdf.page_no_1.py.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2c93416bb9cb0c53496f8b6c1946c9ef4f306decb20642a19a8134a84a3059c4
-size 1843775
+oid sha256:1557401b71d7a00584b7ea5993431a6175b263f402202b60745d5809830a40d6
+size 1843783

groundtruth/parser/fillable_form.pdf.page_no_1.py.json.char.txt

@@ -85,7 +85,7 @@
 (054.00, 607.62) (059.28, 607.62) (059.28, 615.18) (054.00, 615.18)      /T1_1 c	<|special_separator|>
 (059.52, 607.62) (066.62, 607.62) (066.62, 615.18) (059.52, 615.18)      /T1_1 o	<|special_separator|>
 (066.94, 607.62) (073.98, 607.62) (073.98, 615.18) (066.94, 615.18)      /T1_1 n	<|special_separator|>
-(074.26, 607.62) (079.19, 607.62) (079.19, 615.18) (074.26, 615.18)      /T1_1 S	<|special_separator|>
+(074.26, 607.62) (079.19, 607.62) (079.19, 615.18) (074.26, 615.18)      /T1_1 s	<|special_separator|>
 (079.50, 607.62) (082.09, 607.62) (082.09, 615.18) (079.50, 615.18)      /T1_1 i	<|special_separator|>
 (082.42, 607.62) (089.15, 607.62) (089.15, 615.18) (082.42, 615.18)      /T1_1 g	<|special_separator|>
 (089.46, 607.62) (096.50, 607.62) (096.50, 615.18) (089.46, 615.18)      /T1_1 n	<|special_separator|>
@@ -727,7 +727,7 @@
 (140.86, 125.15) (144.19, 125.15) (144.19, 131.98) (140.86, 131.98)      /T1_3 c	<|special_separator|>
 (144.35, 125.15) (146.03, 125.15) (146.03, 131.98) (144.35, 131.98)      /T1_3 t	<|special_separator|>
 (145.95, 125.15) (148.11, 125.15) (148.11, 131.98) (145.95, 131.98)      /T1_3 .	<|special_separator|>
-(415.28, 144.94) (420.28, 144.94) (420.28, 151.76) (415.28, 151.76)      /T1_0 h	<|special_separator|>
+(415.28, 144.94) (420.28, 144.94) (420.28, 151.76) (415.28, 151.76)      /T1_0 H	<|special_separator|>
 (420.46, 144.94) (424.50, 144.94) (424.50, 151.76) (420.46, 151.76)      /T1_0 a	<|special_separator|>
 (424.60, 144.94) (428.30, 144.94) (428.30, 151.76) (424.60, 151.76)      /T1_0 n	<|special_separator|>
 (428.43, 144.94) (432.47, 144.94) (432.47, 151.76) (428.43, 151.76)      /T1_0 d	<|special_separator|>
@@ -758,44 +758,44 @@
 (439.35, 112.94) (441.03, 112.94) (441.03, 119.75) (439.35, 119.75)      /T1_0 t	<|special_separator|>
 (441.21, 112.94) (444.00, 112.94) (444.00, 119.75) (441.21, 119.75)      /T1_0 s	<|special_separator|>
 (050.00, 098.16) (053.92, 098.16) (053.92, 104.97) (050.00, 104.97)      /T1_0 S	<|special_separator|>
-(054.04, 098.16) (055.71, 098.16) (055.71, 104.97) (054.04, 104.97)      /T1_0 I	<|special_separator|>
-(055.91, 098.16) (061.65, 098.16) (061.65, 104.97) (055.91, 104.97)      /T1_0 G	<|special_separator|>
-(061.80, 098.16) (067.64, 098.16) (067.64, 104.97) (061.80, 104.97)      /T1_0 N	<|special_separator|>
-(067.78, 098.16) (072.68, 098.16) (072.68, 104.97) (067.78, 104.97)      /T1_0 A	<|special_separator|>
-(072.46, 098.16) (075.75, 098.16) (075.75, 104.97) (072.46, 104.97)      /T1_0 T	<|special_separator|>
-(075.90, 098.16) (080.70, 098.16) (080.70, 104.97) (075.90, 104.97)      /T1_0 U	<|special_separator|>
-(080.90, 098.16) (084.70, 098.16) (084.70, 104.97) (080.90, 104.97)      /T1_0 R	<|special_separator|>
-(084.79, 098.16) (088.53, 098.16) (088.53, 104.97) (084.79, 104.97)      /T1_0 E	<|special_separator|>
+(054.04, 098.16) (055.71, 098.16) (055.71, 104.97) (054.04, 104.97)      /T1_0 i	<|special_separator|>
+(055.91, 098.16) (061.65, 098.16) (061.65, 104.97) (055.91, 104.97)      /T1_0 g	<|special_separator|>
+(061.80, 098.16) (067.64, 098.16) (067.64, 104.97) (061.80, 104.97)      /T1_0 n	<|special_separator|>
+(067.78, 098.16) (072.68, 098.16) (072.68, 104.97) (067.78, 104.97)      /T1_0 a	<|special_separator|>
+(072.46, 098.16) (075.75, 098.16) (075.75, 104.97) (072.46, 104.97)      /T1_0 t	<|special_separator|>
+(075.90, 098.16) (080.70, 098.16) (080.70, 104.97) (075.90, 104.97)      /T1_0 u	<|special_separator|>
+(080.90, 098.16) (084.70, 098.16) (084.70, 104.97) (080.90, 104.97)      /T1_0 r	<|special_separator|>
+(084.79, 098.16) (088.53, 098.16) (088.53, 104.97) (084.79, 104.97)      /T1_0 e	<|special_separator|>
 (088.57, 098.16) (090.72, 098.16) (090.72, 104.97) (088.57, 104.97)      /T1_0  	<|special_separator|>
 (090.76, 098.16) (095.62, 098.16) (095.62, 104.97) (090.76, 104.97)      /T1_0 &	<|special_separator|>
 (095.66, 098.16) (097.81, 098.16) (097.81, 104.97) (095.66, 104.97)      /T1_0  	<|special_separator|>
 (097.85, 098.16) (101.76, 098.16) (101.76, 104.97) (097.85, 104.97)      /T1_0 S	<|special_separator|>
-(101.89, 098.16) (105.18, 098.16) (105.18, 104.97) (101.89, 104.97)      /T1_0 T	<|special_separator|>
-(104.95, 098.16) (109.85, 098.16) (109.85, 104.97) (104.95, 104.97)      /T1_0 A	<|special_separator|>
-(109.63, 098.16) (112.92, 098.16) (112.92, 104.97) (109.63, 104.97)      /T1_0 T	<|special_separator|>
-(113.07, 098.16) (117.87, 098.16) (117.87, 104.97) (113.07, 104.97)      /T1_0 U	<|special_separator|>
-(117.97, 098.16) (121.88, 098.16) (121.88, 104.97) (117.97, 104.97)      /T1_0 S	<|special_separator|>
+(101.89, 098.16) (105.18, 098.16) (105.18, 104.97) (101.89, 104.97)      /T1_0 t	<|special_separator|>
+(104.95, 098.16) (109.85, 098.16) (109.85, 104.97) (104.95, 104.97)      /T1_0 a	<|special_separator|>
+(109.63, 098.16) (112.92, 098.16) (112.92, 104.97) (109.63, 104.97)      /T1_0 t	<|special_separator|>
+(113.07, 098.16) (117.87, 098.16) (117.87, 104.97) (113.07, 104.97)      /T1_0 u	<|special_separator|>
+(117.97, 098.16) (121.88, 098.16) (121.88, 104.97) (117.97, 104.97)      /T1_0 s	<|special_separator|>
 (121.91, 098.16) (124.07, 098.16) (124.07, 104.97) (121.91, 104.97)      /T1_0  	<|special_separator|>
-(124.11, 098.16) (130.14, 098.16) (130.14, 104.97) (124.11, 104.97)      /T1_0 O	<|special_separator|>
-(130.34, 098.16) (133.69, 098.16) (133.69, 104.97) (130.34, 104.97)      /T1_0 F	<|special_separator|>
+(124.11, 098.16) (130.14, 098.16) (130.14, 104.97) (124.11, 104.97)      /T1_0 o	<|special_separator|>
+(130.34, 098.16) (133.69, 098.16) (133.69, 104.97) (130.34, 104.97)      /T1_0 f	<|special_separator|>
 (133.72, 098.16) (135.88, 098.16) (135.88, 104.97) (133.72, 104.97)      /T1_0  	<|special_separator|>
 (135.91, 098.16) (140.81, 098.16) (140.81, 104.97) (135.91, 104.97)      /T1_0 A	<|special_separator|>
-(140.75, 098.16) (145.55, 098.16) (145.55, 104.97) (140.75, 104.97)      /T1_0 U	<|special_separator|>
-(145.70, 098.16) (148.99, 098.16) (148.99, 104.97) (145.70, 104.97)      /T1_0 T	<|special_separator|>
+(140.75, 098.16) (145.55, 098.16) (145.55, 104.97) (140.75, 104.97)      /T1_0 u	<|special_separator|>
+(145.70, 098.16) (148.99, 098.16) (148.99, 104.97) (145.70, 104.97)      /T1_0 t	<|special_separator|>
 (149.12, 098.16) (154.12, 098.16) (154.12, 104.97) (149.12, 104.97)      /T1_0 h	<|special_separator|>
-(154.31, 098.16) (160.34, 098.16) (160.34, 104.97) (154.31, 104.97)      /T1_0 O	<|special_separator|>
-(160.55, 098.16) (164.34, 098.16) (164.34, 104.97) (160.55, 104.97)      /T1_0 R	<|special_separator|>
-(164.43, 098.16) (166.11, 098.16) (166.11, 104.97) (164.43, 104.97)      /T1_0 I	<|special_separator|>
+(154.31, 098.16) (160.34, 098.16) (160.34, 104.97) (154.31, 104.97)      /T1_0 o	<|special_separator|>
+(160.55, 098.16) (164.34, 098.16) (164.34, 104.97) (160.55, 104.97)      /T1_0 r	<|special_separator|>
+(164.43, 098.16) (166.11, 098.16) (166.11, 104.97) (164.43, 104.97)      /T1_0 i	<|special_separator|>
 (166.22, 098.16) (170.33, 098.16) (170.33, 104.97) (166.22, 104.97)      /T1_0 z	<|special_separator|>
-(170.40, 098.16) (174.14, 098.16) (174.14, 104.97) (170.40, 104.97)      /T1_0 E	<|special_separator|>
-(174.19, 098.16) (178.94, 098.16) (178.94, 104.97) (174.19, 104.97)      /T1_0 D	<|special_separator|>
+(170.40, 098.16) (174.14, 098.16) (174.14, 104.97) (170.40, 104.97)      /T1_0 e	<|special_separator|>
+(174.19, 098.16) (178.94, 098.16) (178.94, 104.97) (174.19, 104.97)      /T1_0 d	<|special_separator|>
 (178.97, 098.16) (181.13, 098.16) (181.13, 104.97) (178.97, 104.97)      /T1_0  	<|special_separator|>
 (181.16, 098.16) (184.68, 098.16) (184.68, 104.97) (181.16, 104.97)      /T1_0 P	<|special_separator|>
-(184.76, 098.16) (188.50, 098.16) (188.50, 104.97) (184.76, 104.97)      /T1_0 E	<|special_separator|>
-(188.55, 098.16) (192.34, 098.16) (192.34, 104.97) (188.55, 104.97)      /T1_0 R	<|special_separator|>
-(192.34, 098.16) (196.25, 098.16) (196.25, 104.97) (192.34, 104.97)      /T1_0 S	<|special_separator|>
-(196.35, 098.16) (202.39, 098.16) (202.39, 104.97) (196.35, 104.97)      /T1_0 O	<|special_separator|>
-(202.58, 098.16) (208.41, 098.16) (208.41, 104.97) (202.58, 104.97)      /T1_0 N	<|special_separator|>
+(184.76, 098.16) (188.50, 098.16) (188.50, 104.97) (184.76, 104.97)      /T1_0 e	<|special_separator|>
+(188.55, 098.16) (192.34, 098.16) (192.34, 104.97) (188.55, 104.97)      /T1_0 r	<|special_separator|>
+(192.34, 098.16) (196.25, 098.16) (196.25, 104.97) (192.34, 104.97)      /T1_0 s	<|special_separator|>
+(196.35, 098.16) (202.39, 098.16) (202.39, 104.97) (196.35, 104.97)      /T1_0 o	<|special_separator|>
+(202.58, 098.16) (208.41, 098.16) (208.41, 104.97) (202.58, 104.97)      /T1_0 n	<|special_separator|>
 (395.50, 096.93) (400.40, 096.93) (400.40, 103.75) (395.50, 103.75)      /T1_0 A	<|special_separator|>
 (400.51, 096.93) (404.55, 096.93) (404.55, 103.75) (400.51, 103.75)      /T1_0 d	<|special_separator|>
 (404.70, 096.93) (408.74, 096.93) (408.74, 103.75) (404.70, 103.75)      /T1_0 d	<|special_separator|>
@@ -826,27 +826,27 @@
 (437.42, 080.93) (441.05, 080.93) (441.05, 087.75) (437.42, 087.75)      /T1_0 e	<|special_separator|>
 (441.21, 080.93) (444.00, 080.93) (444.00, 087.75) (441.21, 087.75)      /T1_0 s	<|special_separator|>
 (050.01, 066.16) (053.53, 066.16) (053.53, 072.97) (050.01, 072.97)      /T1_0 P	<|special_separator|>
-(053.61, 066.16) (057.40, 066.16) (057.40, 072.97) (053.61, 072.97)      /T1_0 R	<|special_separator|>
-(057.49, 066.16) (059.17, 066.16) (059.17, 072.97) (057.49, 072.97)      /T1_0 I	<|special_separator|>
-(059.34, 066.16) (065.18, 066.16) (065.18, 072.97) (059.34, 072.97)      /T1_0 N	<|special_separator|>
-(065.31, 066.16) (068.60, 066.16) (068.60, 072.97) (065.31, 072.97)      /T1_0 T	<|special_separator|>
+(053.61, 066.16) (057.40, 066.16) (057.40, 072.97) (053.61, 072.97)      /T1_0 r	<|special_separator|>
+(057.49, 066.16) (059.17, 066.16) (059.17, 072.97) (057.49, 072.97)      /T1_0 i	<|special_separator|>
+(059.34, 066.16) (065.18, 066.16) (065.18, 072.97) (059.34, 072.97)      /T1_0 n	<|special_separator|>
+(065.31, 066.16) (068.60, 066.16) (068.60, 072.97) (065.31, 072.97)      /T1_0 t	<|special_separator|>
 (068.64, 066.16) (070.79, 066.16) (070.79, 072.97) (068.64, 072.97)      /T1_0  	<|special_separator|>
 (070.83, 066.16) (076.67, 066.16) (076.67, 072.97) (070.83, 072.97)      /T1_0 N	<|special_separator|>
-(076.81, 066.16) (081.71, 066.16) (081.71, 072.97) (076.81, 072.97)      /T1_0 A	<|special_separator|>
-(082.04, 066.16) (088.31, 066.16) (088.31, 072.97) (082.04, 072.97)      /T1_0 M	<|special_separator|>
-(088.52, 066.16) (092.26, 066.16) (092.26, 072.97) (088.52, 072.97)      /T1_0 E	<|special_separator|>
-(092.30, 066.16) (094.45, 066.16) (094.45, 072.97) (092.30, 072.97)      /T1_0  	<|special_separator|>
+(076.81, 066.16) (081.71, 066.16) (081.71, 072.97) (076.81, 072.97)      /T1_0 a	<|special_separator|>
+(082.04, 066.16) (088.31, 066.16) (088.31, 072.97) (082.04, 072.97)      /T1_0 m	<|special_separator|>
+(088.52, 066.16) (092.26, 066.16) (092.26, 072.97) (088.52, 072.97)      /T1_0 e	<|special_separator|>
+(092.30, 066.16) (094.45, 066.16) (094.45, 072.97) (092.30, 072.97)      /T1_0 		<|special_separator|>
 (249.01, 066.16) (253.75, 066.16) (253.75, 072.97) (249.01, 072.97)      /T1_0 D	<|special_separator|>
-(253.65, 066.16) (258.54, 066.16) (258.54, 072.97) (253.65, 072.97)      /T1_0 A	<|special_separator|>
-(258.32, 066.16) (261.61, 066.16) (261.61, 072.97) (258.32, 072.97)      /T1_0 T	<|special_separator|>
-(261.76, 066.16) (265.50, 066.16) (265.50, 072.97) (261.76, 072.97)      /T1_0 E	<|special_separator|>
-(361.26, 065.13) (364.99, 065.13) (364.99, 071.75) (361.26, 071.75)      /T1_1 t	<|special_separator|>
+(253.65, 066.16) (258.54, 066.16) (258.54, 072.97) (253.65, 072.97)      /T1_0 a	<|special_separator|>
+(258.32, 066.16) (261.61, 066.16) (261.61, 072.97) (258.32, 072.97)      /T1_0 t	<|special_separator|>
+(261.76, 066.16) (265.50, 066.16) (265.50, 072.97) (261.76, 072.97)      /T1_0 e	<|special_separator|>
+(361.26, 065.13) (364.99, 065.13) (364.99, 071.75) (361.26, 071.75)      /T1_1 T	<|special_separator|>
 (365.05, 065.13) (371.26, 065.13) (371.26, 071.75) (365.05, 071.75)      /T1_1 o	<|special_separator|>
 (371.33, 065.13) (375.06, 065.13) (375.06, 071.75) (371.33, 071.75)      /T1_1 t	<|special_separator|>
 (374.86, 065.13) (380.33, 065.13) (380.33, 071.75) (374.86, 071.75)      /T1_1 a	<|special_separator|>
 (380.47, 065.13) (383.93, 065.13) (383.93, 071.75) (380.47, 071.75)      /T1_1 l	<|special_separator|>
 (383.97, 065.13) (386.37, 065.13) (386.37, 071.75) (383.97, 071.75)      /T1_1  	<|special_separator|>
-(386.40, 065.13) (388.67, 065.13) (388.67, 071.75) (386.40, 071.75)      /T1_1 i	<|special_separator|>
+(386.40, 065.13) (388.67, 065.13) (388.67, 071.75) (386.40, 071.75)      /T1_1 I	<|special_separator|>
 (388.84, 065.13) (395.01, 065.13) (395.01, 071.75) (388.84, 071.75)      /T1_1 n	<|special_separator|>
 (395.16, 065.13) (400.49, 065.13) (400.49, 071.75) (395.16, 071.75)      /T1_1 v	<|special_separator|>
 (400.42, 065.13) (406.63, 065.13) (406.63, 071.75) (400.42, 071.75)      /T1_1 o	<|special_separator|>
@@ -854,7 +854,7 @@
 (409.25, 065.13) (413.87, 065.13) (413.87, 071.75) (409.25, 071.75)      /T1_1 c	<|special_separator|>
 (414.05, 065.13) (417.90, 065.13) (417.90, 071.75) (414.05, 071.75)      /T1_1 e	<|special_separator|>
 (417.93, 065.13) (420.33, 065.13) (420.33, 071.75) (417.93, 071.75)      /T1_1  	<|special_separator|>
-(420.37, 065.13) (425.69, 065.13) (425.69, 071.75) (420.37, 071.75)      /T1_1 v	<|special_separator|>
+(420.37, 065.13) (425.69, 065.13) (425.69, 071.75) (420.37, 071.75)      /T1_1 V	<|special_separator|>
 (425.37, 065.13) (430.84, 065.13) (430.84, 071.75) (425.37, 071.75)      /T1_1 a	<|special_separator|>
 (430.98, 065.13) (434.44, 065.13) (434.44, 071.75) (430.98, 071.75)      /T1_1 l	<|special_separator|>
 (434.53, 065.13) (439.98, 065.13) (439.98, 071.75) (434.53, 071.75)      /T1_1 u	<|special_separator|>
@@ -927,7 +927,7 @@
 (260.72, 046.04) (263.80, 046.04) (263.80, 052.89) (260.72, 052.89)      /T1_2 c	<|special_separator|>
 (264.07, 046.04) (267.49, 046.04) (267.49, 052.89) (264.07, 052.89)      /T1_2 e	<|special_separator|>
 (267.54, 046.04) (269.57, 046.04) (269.57, 052.89) (267.54, 052.89)      /T1_2 .	<|special_separator|>
-(269.78, 046.02) (272.10, 046.02) (272.10, 053.68) (269.78, 053.68)      /T1_4  	<|special_separator|>
+(269.78, 046.02) (272.10, 046.02) (272.10, 053.68) (269.78, 053.68)      /T1_4 		<|special_separator|>
 (050.00, 731.61) (065.84, 731.61) (065.84, 754.30) (050.00, 754.30)      /T1_1 c	<|special_separator|>
 (066.70, 731.61) (087.99, 731.61) (087.99, 754.30) (066.70, 754.30)      /T1_1 o	<|special_separator|>
 (088.83, 731.61) (112.42, 731.61) (112.42, 754.30) (088.83, 754.30)      /T1_1 m	<|special_separator|>

groundtruth/parser/fillable_form.pdf.page_no_1.py.json.line.txt

@@ -7,7 +7,7 @@
 (139.00, 638.21) (166.35, 638.21) (166.35, 645.03) (139.00, 645.03)      /T1_0 Address:	<|special_separator|>
 (054.00, 618.22) (095.15, 618.22) (095.15, 625.78) (054.00, 625.78)      /T1_1 ultimate	<|special_separator|>
 (139.00, 618.21) (160.36, 618.21) (160.36, 625.03) (139.00, 625.03)      /T1_0 Name:	<|special_separator|>
-(054.00, 607.62) (105.95, 607.62) (105.95, 615.18) (054.00, 615.18)      /T1_1 conSignee	<|special_separator|>
+(054.00, 607.62) (105.95, 607.62) (105.95, 615.18) (054.00, 615.18)      /T1_1 consignee	<|special_separator|>
 (139.00, 598.21) (166.35, 598.21) (166.35, 605.03) (139.00, 605.03)      /T1_0 Address:	<|special_separator|>
 (139.00, 578.21) (155.67, 578.21) (155.67, 585.03) (139.00, 585.03)      /T1_0 Type:	<|special_separator|>
 (139.00, 558.21) (160.36, 558.21) (160.36, 565.03) (139.00, 565.03)      /T1_0 Name:	<|special_separator|>
@@ -40,17 +40,17 @@
 (050.00, 145.16) (275.63, 145.16) (275.63, 151.99) (050.00, 151.99)      /T1_3 It is hereby certified that this invoice shows the actual price of the goods 	<|special_separator|>
 (050.00, 135.15) (273.41, 135.15) (273.41, 141.98) (050.00, 141.98)      /T1_3 described, that no other invoice has been or will be issued, and that all 	<|special_separator|>
 (050.00, 125.15) (148.11, 125.15) (148.11, 131.98) (050.00, 131.98)      /T1_3 particulars are true and correct.	<|special_separator|>
-(415.28, 144.94) (444.01, 144.94) (444.01, 151.76) (415.28, 151.76)      /T1_0 handling	<|special_separator|>
+(415.28, 144.94) (444.01, 144.94) (444.01, 151.76) (415.28, 151.76)      /T1_0 Handling	<|special_separator|>
 (395.45, 128.94) (443.99, 128.94) (443.99, 135.75) (395.45, 135.75)      /T1_0 Insurance Costs	<|special_separator|>
 (423.74, 112.94) (444.00, 112.94) (444.00, 119.75) (423.74, 119.75)      /T1_0 Assists	<|special_separator|>
-(050.00, 098.16) (208.41, 098.16) (208.41, 104.97) (050.00, 104.97)      /T1_0 SIGNATURE & STATUS OF AUThORIzED PERSON	<|special_separator|>
+(050.00, 098.16) (208.41, 098.16) (208.41, 104.97) (050.00, 104.97)      /T1_0 Signature & Status of Authorized Person	<|special_separator|>
 (395.50, 096.93) (444.00, 096.93) (444.00, 103.75) (395.50, 103.75)      /T1_0 Additional Fees	<|special_separator|>
 (398.38, 080.93) (444.00, 080.93) (444.00, 087.75) (398.38, 087.75)      /T1_0 Duties & Taxes	<|special_separator|>
-(050.01, 066.16) (094.45, 066.16) (094.45, 072.97) (050.01, 072.97)      /T1_0 PRINT NAME 	<|special_separator|>
-(249.01, 066.16) (265.50, 066.16) (265.50, 072.97) (249.01, 072.97)      /T1_0 DATE	<|special_separator|>
-(361.26, 065.13) (444.01, 065.13) (444.01, 071.75) (361.26, 071.75)      /T1_1 total invoice value	<|special_separator|>
+(050.01, 066.16) (094.45, 066.16) (094.45, 072.97) (050.01, 072.97)      /T1_0 Print Name		<|special_separator|>
+(249.01, 066.16) (265.50, 066.16) (265.50, 072.97) (249.01, 072.97)      /T1_0 Date	<|special_separator|>
+(361.26, 065.13) (444.01, 065.13) (444.01, 071.75) (361.26, 071.75)      /T1_1 Total Invoice Value	<|special_separator|>
 (048.76, 046.04) (269.57, 046.04) (269.57, 052.89) (048.76, 052.89)      /T1_2 "Incoterms" is a trademark of the International Chamber of Commerce.	<|special_separator|>
-(269.78, 046.02) (272.10, 046.02) (272.10, 053.68) (269.78, 053.68)      /T1_4  	<|special_separator|>
+(269.78, 046.02) (272.10, 046.02) (272.10, 053.68) (269.78, 053.68)      /T1_4 		<|special_separator|>
 (050.00, 731.61) (348.94, 731.61) (348.94, 754.30) (050.00, 754.30)      /T1_1 commercial invoice	<|special_separator|>
 (050.89, 196.90) (285.69, 196.90) (285.69, 202.53) (050.89, 202.53)      /C2_2 These items are controlled by the U.S. government and authorized for export only to the 	<|special_separator|>
 (050.89, 189.70) (279.68, 189.70) (279.68, 195.33) (050.89, 195.33)      /C2_2 country of ultimate destination for use by the ultimate consignee or end-user(s) herein 	<|special_separator|>

groundtruth/parser/fillable_form.pdf.page_no_1.py.json.word.txt

@@ -11,7 +11,7 @@
 (139.00, 638.21) (166.35, 638.21) (166.35, 645.03) (139.00, 645.03)      /T1_0 Address:	<|special_separator|>
 (054.00, 618.22) (095.15, 618.22) (095.15, 625.78) (054.00, 625.78)      /T1_1 ultimate	<|special_separator|>
 (139.00, 618.21) (160.36, 618.21) (160.36, 625.03) (139.00, 625.03)      /T1_0 Name:	<|special_separator|>
-(054.00, 607.62) (105.95, 607.62) (105.95, 615.18) (054.00, 615.18)      /T1_1 conSignee	<|special_separator|>
+(054.00, 607.62) (105.95, 607.62) (105.95, 615.18) (054.00, 615.18)      /T1_1 consignee	<|special_separator|>
 (139.00, 598.21) (166.35, 598.21) (166.35, 605.03) (139.00, 605.03)      /T1_0 Address:	<|special_separator|>
 (139.00, 578.21) (155.67, 578.21) (155.67, 585.03) (139.00, 585.03)      /T1_0 Type:	<|special_separator|>
 (139.00, 558.21) (160.36, 558.21) (160.36, 565.03) (139.00, 565.03)      /T1_0 Name:	<|special_separator|>
@@ -111,27 +111,27 @@
 (097.02, 125.15) (108.52, 125.15) (108.52, 131.98) (097.02, 131.98)      /T1_3 true	<|special_separator|>
 (110.75, 125.15) (122.72, 125.15) (122.72, 131.98) (110.75, 131.98)      /T1_3 and	<|special_separator|>
 (124.95, 125.15) (148.11, 125.15) (148.11, 131.98) (124.95, 131.98)      /T1_3 correct.	<|special_separator|>
-(415.28, 144.94) (444.01, 144.94) (444.01, 151.76) (415.28, 151.76)      /T1_0 handling	<|special_separator|>
+(415.28, 144.94) (444.01, 144.94) (444.01, 151.76) (415.28, 151.76)      /T1_0 Handling	<|special_separator|>
 (395.45, 128.94) (425.09, 128.94) (425.09, 135.75) (395.45, 135.75)      /T1_0 Insurance	<|special_separator|>
 (427.32, 128.94) (443.99, 128.94) (443.99, 135.75) (427.32, 135.75)      /T1_0 Costs	<|special_separator|>
 (423.74, 112.94) (444.00, 112.94) (444.00, 119.75) (423.74, 119.75)      /T1_0 Assists	<|special_separator|>
-(050.00, 098.16) (088.53, 098.16) (088.53, 104.97) (050.00, 104.97)      /T1_0 SIGNATURE	<|special_separator|>
+(050.00, 098.16) (088.53, 098.16) (088.53, 104.97) (050.00, 104.97)      /T1_0 Signature	<|special_separator|>
 (090.76, 098.16) (095.62, 098.16) (095.62, 104.97) (090.76, 104.97)      /T1_0 &	<|special_separator|>
-(097.85, 098.16) (121.88, 098.16) (121.88, 104.97) (097.85, 104.97)      /T1_0 STATUS	<|special_separator|>
-(124.11, 098.16) (133.69, 098.16) (133.69, 104.97) (124.11, 104.97)      /T1_0 OF	<|special_separator|>
-(135.91, 098.16) (178.94, 098.16) (178.94, 104.97) (135.91, 104.97)      /T1_0 AUThORIzED	<|special_separator|>
-(181.16, 098.16) (208.41, 098.16) (208.41, 104.97) (181.16, 104.97)      /T1_0 PERSON	<|special_separator|>
+(097.85, 098.16) (121.88, 098.16) (121.88, 104.97) (097.85, 104.97)      /T1_0 Status	<|special_separator|>
+(124.11, 098.16) (133.69, 098.16) (133.69, 104.97) (124.11, 104.97)      /T1_0 of	<|special_separator|>
+(135.91, 098.16) (178.94, 098.16) (178.94, 104.97) (135.91, 104.97)      /T1_0 Authorized	<|special_separator|>
+(181.16, 098.16) (208.41, 098.16) (208.41, 104.97) (181.16, 104.97)      /T1_0 Person	<|special_separator|>
 (395.50, 096.93) (428.03, 096.93) (428.03, 103.75) (395.50, 103.75)      /T1_0 Additional	<|special_separator|>
 (430.26, 096.93) (444.00, 096.93) (444.00, 103.75) (430.26, 103.75)      /T1_0 Fees	<|special_separator|>
 (398.38, 080.93) (417.29, 080.93) (417.29, 087.75) (398.38, 087.75)      /T1_0 Duties	<|special_separator|>
 (419.52, 080.93) (424.38, 080.93) (424.38, 087.75) (419.52, 087.75)      /T1_0 &	<|special_separator|>
 (426.61, 080.93) (444.00, 080.93) (444.00, 087.75) (426.61, 087.75)      /T1_0 Taxes	<|special_separator|>
-(050.01, 066.16) (068.60, 066.16) (068.60, 072.97) (050.01, 072.97)      /T1_0 PRINT	<|special_separator|>
-(070.83, 066.16) (092.26, 066.16) (092.26, 072.97) (070.83, 072.97)      /T1_0 NAME	<|special_separator|>
-(249.01, 066.16) (265.50, 066.16) (265.50, 072.97) (249.01, 072.97)      /T1_0 DATE	<|special_separator|>
-(361.26, 065.13) (383.93, 065.13) (383.93, 071.75) (361.26, 071.75)      /T1_1 total	<|special_separator|>
-(386.40, 065.13) (417.90, 065.13) (417.90, 071.75) (386.40, 071.75)      /T1_1 invoice	<|special_separator|>
-(420.37, 065.13) (444.01, 065.13) (444.01, 071.75) (420.37, 071.75)      /T1_1 value	<|special_separator|>
+(050.01, 066.16) (068.60, 066.16) (068.60, 072.97) (050.01, 072.97)      /T1_0 Print	<|special_separator|>
+(070.83, 066.16) (092.26, 066.16) (092.26, 072.97) (070.83, 072.97)      /T1_0 Name	<|special_separator|>
+(249.01, 066.16) (265.50, 066.16) (265.50, 072.97) (249.01, 072.97)      /T1_0 Date	<|special_separator|>
+(361.26, 065.13) (383.93, 065.13) (383.93, 071.75) (361.26, 071.75)      /T1_1 Total	<|special_separator|>
+(386.40, 065.13) (417.90, 065.13) (417.90, 071.75) (386.40, 071.75)      /T1_1 Invoice	<|special_separator|>
+(420.37, 065.13) (444.01, 065.13) (444.01, 071.75) (420.37, 071.75)      /T1_1 Value	<|special_separator|>
 (048.76, 046.04) (084.21, 046.04) (084.21, 052.89) (048.76, 052.89)      /T1_2 "Incoterms"	<|special_separator|>
 (086.59, 046.04) (091.14, 046.04) (091.14, 052.89) (086.59, 052.89)      /T1_2 is	<|special_separator|>
 (093.52, 046.04) (097.41, 046.04) (097.41, 052.89) (093.52, 052.89)      /T1_2 a	<|special_separator|>

groundtruth/parser/font_11.pdf.page_no_10.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0cb9f731fcf0f4fcaca652274187c8d4ee583c8ac98640908822eef3f4fe6218
-size 5420138

groundtruth/parser/font_11.pdf.page_no_10.py.json.line.txt

@@ -1,213 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence | Volume 8 | June 2026 | 984-996	<|special_separator|>
-(546.98, 019.82) (562.68, 019.82) (562.68, 026.80) (546.98, 026.80)      /T1_1 993	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 733.65) (114.53, 733.65) (114.53, 741.31) (039.69, 741.31)      /T1_2 to the drug encoder 	<|special_separator|>
-(114.17, 733.66) (121.72, 733.66) (121.72, 741.31) (114.17, 741.31)      /T1_3 W	<|special_separator|>
-(121.26, 732.68) (132.20, 732.68) (132.20, 737.27) (121.26, 737.27)      /T1_2 drug	<|special_separator|>
-(132.19, 733.65) (294.80, 733.65) (294.80, 741.31) (132.19, 741.31)      /T1_2 , which is the same for all drugs in the combi-	<|special_separator|>
-(039.69, 722.90) (136.71, 722.90) (136.71, 730.56) (039.69, 730.56)      /T1_2 nation. The embeddings 	<|special_separator|>
-(136.73, 721.66) (143.58, 721.66) (143.58, 730.43) (136.73, 730.43)      /C0_0 W	<|special_separator|>
-(143.58, 721.10) (155.14, 721.10) (155.14, 726.84) (143.58, 726.84)      /C0_1 drug	<|special_separator|>
-(157.72, 721.65) (163.10, 721.65) (163.10, 730.45) (157.72, 730.45)      /C0_2 =	<|special_separator|>
-(165.17, 721.66) (168.64, 721.66) (168.64, 730.43) (165.17, 730.43)      /C0_0 z	<|special_separator|>
-(168.64, 726.66) (195.38, 726.66) (195.38, 732.39) (168.64, 732.39)      /C0_1 single drug	<|special_separator|>
-(168.64, 719.46) (170.18, 719.46) (170.18, 725.60) (168.64, 725.60)      /C0_0 i	<|special_separator|>
-(196.76, 722.90) (296.23, 722.90) (296.23, 730.56) (196.76, 730.56)      /T1_2  per drug are averaged to 	<|special_separator|>
-(039.69, 712.15) (138.39, 712.15) (138.39, 719.81) (039.69, 719.81)      /T1_2 obtain the final embedding 	<|special_separator|>
-(138.88, 710.92) (142.35, 710.92) (142.35, 719.68) (138.88, 719.68)      /C0_3 z	<|special_separator|>
-(142.35, 715.92) (153.91, 715.92) (153.91, 721.65) (142.35, 721.65)      /C0_4 drug	<|special_separator|>
-(142.35, 708.72) (143.89, 708.72) (143.89, 714.85) (142.35, 714.85)      /C0_3 i	<|special_separator|>
-(153.68, 712.15) (205.91, 712.15) (205.91, 719.81) (153.68, 719.81)      /T1_2 . The resulting 	<|special_separator|>
-(206.38, 710.92) (209.85, 710.92) (209.85, 719.68) (206.38, 719.68)      /C0_5 z	<|special_separator|>
-(209.85, 715.92) (221.42, 715.92) (221.42, 721.65) (209.85, 721.65)      /C0_6 drug	<|special_separator|>
-(209.85, 708.72) (211.39, 708.72) (211.39, 714.85) (209.85, 714.85)      /C0_5 i	<|special_separator|>
-(221.18, 712.15) (294.80, 712.15) (294.80, 719.81) (221.18, 719.81)      /T1_2  has the same dimen-	<|special_separator|>
-(039.69, 701.40) (174.04, 701.40) (174.04, 709.06) (039.69, 709.06)      /T1_2 sions as if one drug were embedded.	<|special_separator|>
-(039.69, 679.94) (083.69, 679.94) (083.69, 687.92) (039.69, 687.92)      /T1_4 Evaluation	<|special_separator|>
-(039.69, 669.15) (296.15, 669.15) (296.15, 676.81) (039.69, 676.81)      /T1_2 We evaluated the Monge Gap and CMonge in different settings. The 	<|special_separator|>
-(039.69, 658.40) (296.27, 658.40) (296.27, 666.06) (039.69, 666.06)      /T1_2 basic Monge models are unaware of conditional contexts, and thus, 	<|special_separator|>
-(039.69, 647.65) (296.27, 647.65) (296.27, 655.31) (039.69, 655.31)      /T1_2 there is a single model per condition. When a Monge model is trained 	<|special_separator|>
-(039.69, 636.90) (296.10, 636.90) (296.10, 644.56) (039.69, 644.56)      /T1_2 on several conditions, it does not receive any conditional information 	<|special_separator|>
-(039.69, 626.15) (204.64, 626.15) (204.64, 633.81) (039.69, 633.81)      /T1_2 and, therefore, treats all conditions as equal.	<|special_separator|>
-(056.69, 615.40) (296.13, 615.40) (296.13, 623.06) (056.69, 623.06)      /T1_2 All IS models were trained on data for all conditions, using an 80/20 	<|special_separator|>
-(039.69, 604.65) (296.12, 604.65) (296.12, 612.31) (039.69, 612.31)      /T1_2 training/testing split for each condition. The evaluation conditions 	<|special_separator|>
-(039.69, 593.90) (296.22, 593.90) (296.22, 601.56) (039.69, 601.56)      /T1_2 were seen during training. For all Monge models, this means that the 	<|special_separator|>
-(039.69, 583.15) (296.14, 583.15) (296.14, 590.81) (039.69, 590.81)      /T1_2 dose and drug cannot be distinguished. Conversely, for all CMonge 	<|special_separator|>
-(039.69, 572.40) (296.25, 572.40) (296.25, 580.06) (039.69, 580.06)      /T1_2 models, the dose and, when applicable, the drug, were encoded and 	<|special_separator|>
-(039.69, 561.65) (192.84, 561.65) (192.84, 569.31) (039.69, 569.31)      /T1_2 this information was given to the models.	<|special_separator|>
-(056.69, 550.90) (296.23, 550.90) (296.23, 558.56) (056.69, 558.56)      /T1_2 In the OOS setting, the Monge and CMonge models were evaluated 	<|special_separator|>
-(039.69, 540.15) (296.10, 540.15) (296.10, 547.81) (039.69, 547.81)      /T1_2 on held-out conditions and trained on all other conditions. Unless 	<|special_separator|>
-(039.69, 529.40) (289.68, 529.40) (289.68, 537.06) (039.69, 537.06)      /T1_2 specified, we employed a leave-one-out setting where we trained 	<|special_separator|>
-(289.88, 529.41) (294.85, 529.41) (294.85, 537.06) (289.88, 537.06)      /T1_3 n	<|special_separator|>
-(294.85, 529.40) (296.32, 529.40) (296.32, 537.06) (294.85, 537.06)      /T1_2  	<|special_separator|>
-(039.69, 518.65) (082.47, 518.65) (082.47, 526.31) (039.69, 526.31)      /T1_2 models for 	<|special_separator|>
-(082.63, 518.66) (087.56, 518.66) (087.56, 526.31) (082.63, 526.31)      /T1_3 n	<|special_separator|>
-(087.56, 518.65) (296.32, 518.65) (296.32, 526.31) (087.56, 526.31)      /T1_2  held-out conditions, always leaving one condition out. 	<|special_separator|>
-(039.69, 507.90) (296.25, 507.90) (296.25, 515.56) (039.69, 515.56)      /T1_2 As in the IS setting, the Monge models do not distinguish conditions in 	<|special_separator|>
-(039.69, 497.15) (296.31, 497.15) (296.31, 504.81) (039.69, 504.81)      /T1_2 the training or evaluation setting, whereas CMonge can be conditioned 	<|special_separator|>
-(039.69, 486.40) (157.24, 486.40) (157.24, 494.06) (039.69, 494.06)      /T1_2 on dose alone or drug and dose.	<|special_separator|>
-(056.69, 475.65) (296.06, 475.65) (296.06, 483.31) (056.69, 483.31)      /T1_2 First, we tested the ability of a model to condition on the dose and 	<|special_separator|>
-(039.69, 464.90) (296.20, 464.90) (296.20, 472.56) (039.69, 472.56)      /T1_2 to generalize to unseen doses. Therefore, the CMonge models were 	<|special_separator|>
-(039.69, 454.15) (296.27, 454.15) (296.27, 461.81) (039.69, 461.81)      /T1_2 conditioned on dose (Dose) in both an IS setting and an OOS setting. 	<|special_separator|>
-(039.69, 443.40) (296.25, 443.40) (296.25, 451.06) (039.69, 451.06)      /T1_2 Then, we tested the ability of the model to condition on drugs and 	<|special_separator|>
-(039.69, 432.65) (296.21, 432.65) (296.21, 440.31) (039.69, 440.31)      /T1_2 generalize to unseen drugs by holding out all doses of one drug during 	<|special_separator|>
-(039.69, 421.90) (296.20, 421.90) (296.20, 429.56) (039.69, 429.56)      /T1_2 training. The CMonge models were conditioned on both drug and dose 	<|special_separator|>
-(039.69, 411.15) (296.24, 411.15) (296.24, 418.81) (039.69, 418.81)      /T1_2 (DrugDose) in an IS setting and an OOS setting. Last, we investigated 	<|special_separator|>
-(039.69, 400.40) (296.20, 400.40) (296.20, 408.06) (039.69, 408.06)      /T1_2 model performance on conditioning on DrugDose with more training 	<|special_separator|>
-(039.69, 389.65) (296.29, 389.65) (296.29, 397.31) (039.69, 397.31)      /T1_2 drugs, both in an IS setting and an OOS setting. In the OOS setting, we 	<|special_separator|>
-(039.69, 378.90) (294.75, 378.90) (294.75, 386.56) (039.69, 386.56)      /T1_2 increased the OOS conditions by leaving nine drugs out instead of one.	<|special_separator|>
-(056.69, 368.15) (121.21, 368.15) (121.21, 375.81) (056.69, 375.81)      /T1_2 We evaluated the 	<|special_separator|>
-(120.84, 368.16) (125.98, 368.16) (125.98, 375.81) (120.84, 375.81)      /T1_3 R	<|special_separator|>
-(126.01, 372.34) (128.69, 372.34) (128.69, 376.93) (126.01, 376.93)      /T1_2 2	<|special_separator|>
-(128.68, 368.15) (296.15, 368.15) (296.15, 375.81) (128.68, 375.81)      /T1_2  metric between the perturbed and predicted 	<|special_separator|>
-(039.69, 357.40) (296.26, 357.40) (296.26, 365.06) (039.69, 365.06)      /T1_2 feature means, the Wasserstein distance (equation (2)) and the MMD. 	<|special_separator|>
-(039.69, 346.65) (296.24, 346.65) (296.24, 354.31) (039.69, 354.31)      /T1_2 All metrics were calculated on batches of observations sampled from 	<|special_separator|>
-(039.69, 335.90) (296.25, 335.90) (296.25, 343.56) (039.69, 343.56)      /T1_2 the test set and we report the mean across these batches. Specifically, 	<|special_separator|>
-(039.69, 325.15) (296.28, 325.15) (296.28, 332.81) (039.69, 332.81)      /T1_2 for each condition, we always performed nine independent evaluations 	<|special_separator|>
-(039.69, 314.40) (296.11, 314.40) (296.11, 322.06) (039.69, 322.06)      /T1_2 by randomly sampling a set of control cells. Thus, all reported metrics 	<|special_separator|>
-(039.69, 303.65) (042.75, 303.65) (042.75, 311.31) (039.69, 311.31)      /T1_2 (	<|special_separator|>
-(042.78, 303.66) (047.98, 303.66) (047.98, 311.31) (042.78, 311.31)      /T1_3 R	<|special_separator|>
-(048.01, 307.84) (050.71, 307.84) (050.71, 312.43) (048.01, 312.43)      /T1_2 2	<|special_separator|>
-(050.73, 303.65) (296.30, 303.65) (296.30, 311.31) (050.73, 311.31)      /T1_2 , MMD and Wasserstein distance) were computed per condition, 	<|special_separator|>
-(039.69, 292.90) (296.28, 292.90) (296.28, 300.56) (039.69, 300.56)      /T1_2 including this bootstrapping procedure, and then averaged across all 	<|special_separator|>
-(039.69, 282.15) (296.32, 282.15) (296.32, 289.81) (039.69, 289.81)      /T1_2 conditions (if any). The experiments used the same hyperparameters, 	<|special_separator|>
-(039.69, 271.40) (296.24, 271.40) (296.24, 279.06) (039.69, 279.06)      /T1_2 except for the number of optimization steps and the latent encoding 	<|special_separator|>
-(039.69, 260.65) (296.13, 260.65) (296.13, 268.31) (039.69, 268.31)      /T1_2 of the features and context (for more details, see 'Hyperparameters 	<|special_separator|>
-(039.69, 249.90) (296.24, 249.90) (296.24, 257.56) (039.69, 257.56)      /T1_2 and model sizes' in Methods). For the SciPlex dataset, we trained and 	<|special_separator|>
-(039.69, 239.15) (203.35, 239.15) (203.35, 246.81) (039.69, 246.81)      /T1_2 evaluated the CM in the following scenarios:	<|special_separator|>
-(039.69, 224.40) (050.67, 224.40) (050.67, 232.06) (039.69, 232.06)      /T1_2   (1) 	<|special_separator|>
-(056.13, 224.40) (290.12, 224.40) (290.12, 232.06) (056.13, 232.06)      /T1_2 Monge: As a hypothetical upper bound on performance, we fit-	<|special_separator|>
-(056.13, 213.65) (295.29, 213.65) (295.29, 221.31) (056.13, 221.31)      /T1_2 ted separate Monge Gap models, one per drug-dose pair. These 	<|special_separator|>
-(056.13, 202.90) (251.43, 202.90) (251.43, 210.56) (056.13, 210.56)      /T1_2 models do not have any context (Uscidda and Cuturi	<|special_separator|>
-(251.43, 207.09) (256.74, 207.09) (256.74, 211.68) (251.43, 211.68)      /T1_2 32	<|special_separator|>
-(256.74, 202.90) (262.01, 202.90) (262.01, 210.56) (256.74, 210.56)      /T1_2 ).	<|special_separator|>
-(039.68, 192.15) (051.70, 192.15) (051.70, 199.81) (039.68, 199.81)      /T1_2   (2) 	<|special_separator|>
-(056.13, 192.15) (280.72, 192.15) (280.72, 199.81) (056.13, 199.81)      /T1_2 Monge-Drug and Monge-DrugDose: To motivate the contex-	<|special_separator|>
-(056.13, 181.40) (289.05, 181.40) (289.05, 189.06) (056.13, 189.06)      /T1_2 tual settings, we fitted a Monge Gap model that was trained on 	<|special_separator|>
-(056.13, 170.65) (294.13, 170.65) (294.13, 178.31) (056.13, 178.31)      /T1_2 several conditions but was unaware of conditional information, 	<|special_separator|>
-(056.13, 159.90) (260.91, 159.90) (260.91, 167.56) (056.13, 167.56)      /T1_2 and we evaluated it on conditions seen during training.	<|special_separator|>
-(053.58, 146.31) (058.05, 146.31) (058.05, 153.97) (053.58, 153.97)      /T1_2 • 	<|special_separator|>
-(067.47, 146.31) (292.23, 146.31) (292.23, 153.97) (067.47, 153.97)      /T1_2 Monge-Dose-IS: a homogeneous model for each drug, using 	<|special_separator|>
-(067.47, 135.56) (158.14, 135.56) (158.14, 143.22) (067.47, 143.22)      /T1_2 data from all four doses.	<|special_separator|>
-(053.58, 124.81) (058.05, 124.81) (058.05, 132.47) (053.58, 132.47)      /T1_2 • 	<|special_separator|>
-(067.47, 124.81) (289.59, 124.81) (289.59, 132.47) (067.47, 132.47)      /T1_2 Monge-DrugDose-IS: one model on all conditions (all drug-	<|special_separator|>
-(067.47, 114.06) (111.06, 114.06) (111.06, 121.72) (067.47, 121.72)      /T1_2 dose pairs).	<|special_separator|>
-(053.58, 103.31) (058.05, 103.31) (058.05, 110.97) (053.58, 110.97)      /T1_2 • 	<|special_separator|>
-(067.47, 103.31) (276.71, 103.31) (276.71, 110.97) (067.47, 110.97)      /T1_2 Monge-Dose-OOS: a model for each drug, with different 	<|special_separator|>
-(067.47, 092.56) (178.91, 092.56) (178.91, 100.22) (067.47, 100.22)      /T1_2 doses left out during training.	<|special_separator|>
-(053.58, 081.82) (058.05, 081.82) (058.05, 089.47) (053.58, 089.47)      /T1_2 • 	<|special_separator|>
-(067.47, 081.82) (282.25, 081.82) (282.25, 089.47) (067.47, 089.47)      /T1_2 Monge-DrugDose-OOS: a model trained on all conditions 	<|special_separator|>
-(067.47, 071.07) (206.69, 071.07) (206.69, 078.72) (067.47, 078.72)      /T1_2 but the doses of the held-out drug(s).	<|special_separator|>
-(039.69, 057.31) (051.42, 057.31) (051.42, 064.97) (039.69, 064.97)      /T1_2   (3) 	<|special_separator|>
-(056.14, 057.31) (289.00, 057.31) (289.00, 064.97) (056.14, 064.97)      /T1_2 CMonge-Dose-IS and CMonge-Dose-OOS: We fitted conditional 	<|special_separator|>
-(056.14, 046.56) (296.15, 046.56) (296.15, 054.22) (056.14, 054.22)      /T1_2 models for each drug with the scalar dose as context. The IS setting 	<|special_separator|>
-(322.59, 733.67) (556.38, 733.67) (556.38, 741.33) (322.59, 741.33)      /T1_2 saw all doses during training. For the OOS setting, we left out differ-	<|special_separator|>
-(322.59, 722.92) (561.16, 722.92) (561.16, 730.58) (322.59, 730.58)      /T1_2 ent doses during training, thus creating interpolation and extrapola-	<|special_separator|>
-(322.59, 712.17) (368.64, 712.17) (368.64, 719.83) (322.59, 719.83)      /T1_2 tion settings.	<|special_separator|>
-(306.15, 701.42) (554.28, 701.42) (554.28, 709.08) (306.15, 709.08)      /T1_2   (4)  CMonge-DrugDose-RDKit and CMonge-DrugDose-MoA: A sin-	<|special_separator|>
-(322.59, 690.67) (553.62, 690.67) (553.62, 698.33) (322.59, 698.33)      /T1_2 gle model was fitted to all data, conditioned on drug and dose 	<|special_separator|>
-(322.59, 679.92) (543.96, 679.92) (543.96, 687.58) (322.59, 687.58)      /T1_2 context. To encode the drug, we compared the fingerprints 	<|special_separator|>
-(322.59, 669.17) (476.46, 669.17) (476.46, 676.83) (322.59, 676.83)      /T1_2 (RDKit) to a data-driven approach (MoA).	<|special_separator|>
-(320.04, 651.59) (324.52, 651.59) (324.52, 659.25) (320.04, 659.25)      /T1_2 • 	<|special_separator|>
-(333.94, 651.59) (545.99, 651.59) (545.99, 659.25) (333.94, 659.25)      /T1_2 CMonge-DrugDose-x-IS: All conditions were seen during 	<|special_separator|>
-(333.94, 640.84) (366.31, 640.84) (366.31, 648.50) (333.94, 648.50)      /T1_2 training.	<|special_separator|>
-(320.04, 630.09) (324.52, 630.09) (324.52, 637.75) (320.04, 637.75)      /T1_2 • 	<|special_separator|>
-(333.94, 630.09) (551.76, 630.09) (551.76, 637.75) (333.94, 637.75)      /T1_2 CMonge-DrugDose-x-OOS: All doses of one or more drugs 	<|special_separator|>
-(333.94, 619.34) (485.32, 619.34) (485.32, 627.00) (333.94, 627.00)      /T1_2 were held during training for evaluation.	<|special_separator|>
-(323.16, 604.66) (561.28, 604.66) (561.28, 612.32) (323.16, 612.32)      /T1_2 For the conditional Monge 4i experiments, combinatorial thera-	<|special_separator|>
-(306.14, 593.90) (562.73, 593.90) (562.73, 601.56) (306.14, 601.56)      /T1_2 pies (conditions with two or three drugs) were handled similarly for 	<|special_separator|>
-(306.14, 583.15) (562.72, 583.15) (562.72, 590.81) (306.14, 590.81)      /T1_2 the RDKit and the MoA embeddings. For each drug in the condition, 	<|special_separator|>
-(306.14, 572.40) (562.74, 572.40) (562.74, 580.06) (306.14, 580.06)      /T1_2 the embedding was computed for the MoA based on the single-drug 	<|special_separator|>
-(306.14, 561.65) (562.68, 561.65) (562.68, 569.31) (306.14, 569.31)      /T1_2 condition, if possible. Then, the embeddings from all drugs in the 	<|special_separator|>
-(306.14, 550.90) (562.64, 550.90) (562.64, 558.56) (306.14, 558.56)      /T1_2 condition were passed through the same dense layer, with the same 	<|special_separator|>
-(306.14, 540.15) (562.84, 540.15) (562.84, 547.81) (306.14, 547.81)      /T1_2 parameters for each single-drug embedding. We left out the condition 	<|special_separator|>
-(306.14, 529.40) (562.72, 529.40) (562.72, 537.06) (306.14, 537.06)      /T1_2 vemurafenib-cobimetinib, as neither drug was measured in isolation 	<|special_separator|>
-(306.14, 518.65) (562.67, 518.65) (562.67, 526.31) (306.14, 526.31)      /T1_2 and, therefore, no MoA embedding was possible for either single drug. 	<|special_separator|>
-(306.14, 507.90) (562.74, 507.90) (562.74, 515.56) (306.14, 515.56)      /T1_2 We also left out pomalidomide-carfilzomib-dexamethasone, as only 	<|special_separator|>
-(306.14, 497.15) (492.33, 497.15) (492.33, 504.81) (306.14, 504.81)      /T1_2 dexamethasone was present as a single treatment.	<|special_separator|>
-(306.14, 475.69) (341.89, 475.69) (341.89, 483.67) (306.14, 483.67)      /T1_4 Datasets	<|special_separator|>
-(306.14, 464.90) (562.52, 464.90) (562.52, 472.56) (306.14, 472.56)      /T1_2 We used two datasets to train and evaluate our models. First, the SciPlex 	<|special_separator|>
-(306.14, 454.15) (333.38, 454.15) (333.38, 461.81) (306.14, 461.81)      /T1_2 dataset	<|special_separator|>
-(333.35, 458.34) (336.01, 458.34) (336.01, 462.93) (333.35, 462.93)      /T1_2 3	<|special_separator|>
-(335.99, 454.15) (562.63, 454.15) (562.63, 461.81) (335.99, 461.81)      /T1_2  contains single-cell profiles from three human cancer cell lines 	<|special_separator|>
-(306.14, 443.40) (561.26, 443.40) (561.26, 451.06) (306.14, 451.06)      /T1_2 (A549, K562 and MCF7) that were exposed to 188 compounds, compris-	<|special_separator|>
-(306.14, 432.65) (562.57, 432.65) (562.57, 440.31) (306.14, 440.31)      /T1_2 ing 187 drugs that were administered at four different doses (10 nM, 	<|special_separator|>
-(306.14, 421.90) (561.26, 421.90) (561.26, 429.56) (306.14, 429.56)      /T1_2 100 nM, 1,000 nM and 10,000 nM) and a control solution. We per-	<|special_separator|>
-(306.14, 411.15) (562.56, 411.15) (562.56, 418.81) (306.14, 418.81)      /T1_2 formed experiments on a selection of nine different drugs, as in Uscidda 	<|special_separator|>
-(306.14, 400.40) (346.31, 400.40) (346.31, 408.06) (306.14, 408.06)      /T1_2 and Cuturi	<|special_separator|>
-(346.31, 404.59) (351.63, 404.59) (351.63, 409.18) (346.31, 409.18)      /T1_2 32	<|special_separator|>
-(351.63, 400.40) (554.86, 400.40) (554.86, 408.06) (351.63, 408.06)      /T1_2 , and on the full range of compounds, as in Hetzel et al.	<|special_separator|>
-(554.89, 404.59) (559.03, 404.59) (559.03, 409.18) (554.89, 409.18)      /T1_2 11	<|special_separator|>
-(559.03, 400.40) (562.71, 400.40) (562.71, 408.06) (559.03, 408.06)      /T1_2 . 	<|special_separator|>
-(306.15, 389.65) (498.96, 389.65) (498.96, 397.31) (306.15, 397.31)      /T1_2 We used the preprocessed data from Lotfollahi et al.	<|special_separator|>
-(498.88, 393.84) (501.79, 393.84) (501.79, 398.43) (498.88, 398.43)      /T1_2 8	<|special_separator|>
-(501.74, 389.65) (562.71, 389.65) (562.71, 397.31) (501.74, 397.31)      /T1_2 , which includes 	<|special_separator|>
-(306.15, 378.90) (513.45, 378.90) (513.45, 386.56) (306.15, 386.56)      /T1_2 library size normalization, cell and gene filtering, and a 	<|special_separator|>
-(513.58, 379.22) (537.83, 379.22) (537.83, 385.33) (513.58, 385.33)      /T1_5 log1p	<|special_separator|>
-(537.71, 378.90) (561.26, 378.90) (561.26, 386.56) (537.71, 386.56)      /T1_2  trans-	<|special_separator|>
-(306.14, 368.15) (562.61, 368.15) (562.61, 375.81) (306.14, 375.81)      /T1_2 formation. The dataset consists of 762,039 single-cell measurements, 	<|special_separator|>
-(306.14, 357.40) (562.68, 357.40) (562.68, 365.06) (306.14, 365.06)      /T1_2 out of which 17,565 belong to the control population and, on average, 	<|special_separator|>
-(306.14, 346.65) (562.75, 346.65) (562.75, 354.31) (306.14, 354.31)      /T1_2 4,032 observations of each drug and drug dose condition. During 	<|special_separator|>
-(306.14, 335.90) (562.71, 335.90) (562.71, 343.56) (306.14, 343.56)      /T1_2 training and evaluation, we considered only the 1,000 highly variable 	<|special_separator|>
-(306.14, 325.15) (436.15, 325.15) (436.15, 332.81) (306.14, 332.81)      /T1_2 genes computed by Lotfollahi et al.	<|special_separator|>
-(436.06, 329.34) (438.97, 329.34) (438.97, 333.93) (436.06, 333.93)      /T1_2 8	<|special_separator|>
-(438.97, 325.15) (562.72, 325.15) (562.72, 332.81) (438.97, 332.81)      /T1_2 . As many genes were unaffected, 	<|special_separator|>
-(306.14, 314.40) (562.67, 314.40) (562.67, 322.06) (306.14, 322.06)      /T1_2 instead of evaluating in the 1,000-dimensional gene space, we used 	<|special_separator|>
-(306.14, 303.65) (562.59, 303.65) (562.59, 311.31) (306.14, 311.31)      /T1_2 only the top 50 differentially expressed marker genes obtained through 	<|special_separator|>
-(306.14, 292.90) (355.36, 292.90) (355.36, 300.56) (306.14, 300.56)      /T1_2 gene ranking	<|special_separator|>
-(355.34, 297.09) (360.89, 297.09) (360.89, 301.68) (355.34, 301.68)      /T1_2 62	<|special_separator|>
-(360.90, 292.90) (562.82, 292.90) (562.82, 300.56) (360.90, 300.56)      /T1_2 . To facilitate training, we reduced the dimensionality 	<|special_separator|>
-(306.14, 282.15) (540.40, 282.15) (540.40, 289.81) (306.14, 289.81)      /T1_2 of the 1,000-dimensional gene expression following Bunne et al.	<|special_separator|>
-(540.54, 286.34) (546.48, 286.34) (546.48, 290.93) (540.54, 290.93)      /T1_2 30	<|special_separator|>
-(546.38, 282.15) (562.70, 282.15) (562.70, 289.81) (546.38, 289.81)      /T1_2 . We 	<|special_separator|>
-(306.14, 271.40) (443.69, 271.40) (443.69, 279.06) (306.14, 279.06)      /T1_2 encoded gene expression data into a 	<|special_separator|>
-(443.64, 271.41) (447.98, 271.41) (447.98, 279.06) (443.64, 279.06)      /T1_3 k	<|special_separator|>
-(447.98, 271.40) (562.71, 271.40) (562.71, 279.06) (447.98, 279.06)      /T1_2  = 50-dimensional latent space 	<|special_separator|>
-(306.14, 260.65) (493.35, 260.65) (493.35, 268.31) (306.14, 268.31)      /T1_2 by training a vanilla autoencoder with an encoder 	<|special_separator|>
-(494.26, 259.42) (502.00, 258.72) (502.00, 264.86) (494.26, 268.18)      /C0_7 Eϕ	<|special_separator|>
-(504.58, 259.41) (516.00, 259.41) (516.00, 268.20) (504.58, 268.20)      /C0_8 ∶ ℝ	<|special_separator|>
-(516.01, 262.98) (519.07, 262.98) (519.07, 269.12) (516.01, 269.12)      /C0_7 d	<|special_separator|>
-(521.65, 259.41) (536.25, 259.41) (536.25, 268.20) (521.65, 268.20)      /C0_8 →ℝ	<|special_separator|>
-(536.25, 262.98) (539.00, 262.98) (539.00, 269.12) (536.25, 269.12)      /C0_7 k	<|special_separator|>
-(539.50, 260.65) (562.71, 260.65) (562.71, 268.31) (539.50, 268.31)      /T1_2  and a 	<|special_separator|>
-(306.14, 249.90) (338.12, 249.90) (338.12, 257.56) (306.14, 257.56)      /T1_2 decoder 	<|special_separator|>
-(338.39, 248.66) (346.59, 247.97) (346.59, 254.11) (338.39, 257.43)      /C0_9 Dθ	<|special_separator|>
-(349.16, 248.65) (360.58, 248.65) (360.58, 257.45) (349.16, 257.45)     /C0_10 ∶ ℝ	<|special_separator|>
-(360.59, 252.23) (363.34, 252.23) (363.34, 258.36) (360.59, 258.36)      /C0_9 k	<|special_separator|>
-(365.92, 248.65) (380.52, 248.65) (380.52, 257.45) (365.92, 257.45)     /C0_10 →ℝ	<|special_separator|>
-(380.52, 252.23) (383.59, 252.23) (383.59, 258.36) (380.52, 258.36)      /C0_9 d	<|special_separator|>
-(383.06, 249.90) (405.35, 249.90) (405.35, 257.56) (383.06, 257.56)      /T1_2 . Both 	<|special_separator|>
-(404.92, 249.91) (412.67, 248.93) (412.67, 253.52) (404.92, 257.56)      /T1_3 Eφ	<|special_separator|>
-(412.64, 249.90) (428.90, 249.90) (428.90, 257.56) (412.64, 257.56)      /T1_2  and 	<|special_separator|>
-(428.47, 249.91) (436.83, 248.93) (436.83, 253.52) (428.47, 257.56)      /T1_3 Dθ	<|special_separator|>
-(436.80, 249.90) (562.71, 249.90) (562.71, 257.56) (436.80, 257.56)      /T1_2  were parameterized by multi-layer 	<|special_separator|>
-(306.14, 239.15) (562.65, 239.15) (562.65, 246.81) (306.14, 246.81)      /T1_2 perceptrons.  The  entire  autoencoder  was  optimized  using  a 	<|special_separator|>
-(306.14, 228.40) (561.26, 228.40) (561.26, 236.06) (306.14, 236.06)      /T1_2 mean-squared error reconstruction loss. This autoencoder was pre-	<|special_separator|>
-(306.14, 217.65) (562.68, 217.65) (562.68, 225.31) (306.14, 225.31)      /T1_2 trained once on one batch of each condition, and it was then frozen 	<|special_separator|>
-(306.14, 206.90) (562.72, 206.90) (562.72, 214.56) (306.14, 214.56)      /T1_2 during CMonge training and evaluation. OT learning occurred in the 	<|special_separator|>
-(306.14, 196.15) (513.12, 196.15) (513.12, 203.81) (306.14, 203.81)      /T1_2 latent space of the autoencoder. The reported metrics (	<|special_separator|>
-(513.12, 196.16) (518.27, 196.16) (518.27, 203.81) (513.12, 203.81)      /T1_3 R	<|special_separator|>
-(518.24, 200.34) (520.92, 200.34) (520.92, 204.93) (518.24, 204.93)      /T1_2 2	<|special_separator|>
-(520.92, 196.15) (562.72, 196.15) (562.72, 203.81) (520.92, 203.81)      /T1_2 , MMD and 	<|special_separator|>
-(306.14, 185.40) (562.89, 185.40) (562.89, 193.06) (306.14, 193.06)      /T1_2 Wasserstein) were calculated in the original gene expression space 	<|special_separator|>
-(306.14, 174.65) (535.02, 174.65) (535.02, 182.31) (306.14, 182.31)      /T1_2 (between measured gene expression and decoded transport).	<|special_separator|>
-(323.15, 163.90) (561.18, 163.90) (561.18, 171.56) (323.15, 171.56)      /T1_2 Second, the 4i dataset contains multiplexed protein measure-	<|special_separator|>
-(306.14, 153.15) (561.26, 153.15) (561.26, 160.81) (306.14, 160.81)      /T1_2 ments of 40 different proteins measured in 97,748 cells (10,995 con-	<|special_separator|>
-(306.14, 142.40) (562.79, 142.40) (562.79, 150.06) (306.14, 150.06)      /T1_2 trols) from two lines of melanoma tumours treated with one of 35 	<|special_separator|>
-(306.14, 131.65) (470.98, 131.65) (470.98, 139.31) (306.14, 139.31)      /T1_2 cancer therapies, each involving ~2,500 cells	<|special_separator|>
-(470.97, 135.84) (476.63, 135.84) (476.63, 140.43) (470.97, 140.43)      /T1_2 36	<|special_separator|>
-(476.55, 131.65) (561.27, 131.65) (561.27, 139.31) (476.55, 139.31)      /T1_2 . We obtained preproc-	<|special_separator|>
-(306.14, 120.90) (562.72, 120.90) (562.72, 128.56) (306.14, 128.56)      /T1_2 essed data from ref. 33, resulting in 48 features for cellular marker 	<|special_separator|>
-(306.14, 110.15) (562.83, 110.15) (562.83, 117.81) (306.14, 117.81)      /T1_2 expression and cell shape. We excluded two of the eight combinatorial 	<|special_separator|>
-(306.14, 099.40) (562.70, 099.40) (562.70, 107.06) (306.14, 107.06)      /T1_2 treatments because they did not occur as a single treatment, as we 	<|special_separator|>
-(306.14, 088.65) (519.60, 088.65) (519.60, 096.31) (306.14, 096.31)      /T1_2 based the MoA drug embedding on the single treatments.	<|special_separator|>
-(306.14, 067.19) (445.25, 067.19) (445.25, 075.17) (306.14, 075.17)      /T1_4 Hyperparameters and model sizes	<|special_separator|>
-(306.14, 056.40) (500.39, 056.40) (500.39, 064.06) (306.14, 064.06)      /T1_2 In all experiments, we used the AdamW optimizer	<|special_separator|>
-(500.41, 060.59) (506.17, 060.59) (506.17, 065.18) (500.41, 065.18)      /T1_2 63	<|special_separator|>
-(506.22, 056.40) (562.77, 056.40) (562.77, 064.06) (506.22, 064.06)      /T1_2  with an initial 	<|special_separator|>
-(306.14, 045.65) (373.35, 045.65) (373.35, 053.31) (306.14, 053.31)      /T1_2 learning rate of 10	<|special_separator|>
-(373.34, 049.84) (379.06, 049.84) (379.06, 054.43) (373.34, 054.43)      /T1_2 -4	<|special_separator|>
-(379.06, 045.65) (510.43, 045.65) (510.43, 053.31) (379.06, 053.31)      /T1_2  and weight decay regularization 10	<|special_separator|>
-(510.42, 049.84) (515.71, 049.84) (515.71, 054.43) (510.42, 054.43)      /T1_2 -5	<|special_separator|>
-(515.58, 045.65) (562.72, 045.65) (562.72, 053.31) (515.58, 053.31)      /T1_2 . The Monge 

groundtruth/parser/font_11.pdf.page_no_10.py.json.word.txt

@@ -1,541 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 NatureMachineIntelligence|Volume8|June2026|984-996	<|special_separator|>
-(546.98, 019.82) (562.68, 019.82) (562.68, 026.80) (546.98, 026.80)      /T1_1 993	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 733.65) (113.08, 733.65) (113.08, 741.31) (039.69, 741.31)      /T1_2 tothedrugencoder	<|special_separator|>
-(114.17, 733.66) (121.72, 733.66) (121.72, 741.31) (114.17, 741.31)      /T1_3 W	<|special_separator|>
-(121.26, 732.68) (132.20, 732.68) (132.20, 737.27) (121.26, 737.27)      /T1_2 drug	<|special_separator|>
-(132.19, 733.65) (294.80, 733.65) (294.80, 741.31) (132.19, 741.31)      /T1_2 ,whichisthesameforalldrugsinthecombi-	<|special_separator|>
-(039.69, 722.90) (067.59, 722.90) (067.59, 730.56) (039.69, 730.56)      /T1_2 nation.	<|special_separator|>
-(069.46, 722.90) (084.39, 722.90) (084.39, 730.56) (069.46, 730.56)      /T1_2 The	<|special_separator|>
-(086.26, 722.90) (135.15, 722.90) (135.15, 730.56) (086.26, 730.56)      /T1_2 embeddings	<|special_separator|>
-(136.73, 721.66) (143.58, 721.66) (143.58, 730.43) (136.73, 730.43)      /C0_0 W	<|special_separator|>
-(143.58, 721.10) (155.14, 721.10) (155.14, 726.84) (143.58, 726.84)      /C0_1 drug	<|special_separator|>
-(157.72, 721.65) (163.10, 721.65) (163.10, 730.45) (157.72, 730.45)      /C0_2 =	<|special_separator|>
-(165.17, 721.66) (168.64, 721.66) (168.64, 730.43) (165.17, 730.43)      /C0_0 z	<|special_separator|>
-(168.64, 726.66) (195.38, 726.66) (195.38, 732.39) (168.64, 732.39)      /C0_1 singledrug	<|special_separator|>
-(168.64, 719.46) (170.18, 719.46) (170.18, 725.60) (168.64, 725.60)      /C0_0 i	<|special_separator|>
-(198.55, 722.90) (211.88, 722.90) (211.88, 730.56) (198.55, 730.56)      /T1_2 per	<|special_separator|>
-(213.75, 722.90) (232.61, 722.90) (232.61, 730.56) (213.75, 730.56)      /T1_2 drug	<|special_separator|>
-(234.48, 722.90) (246.85, 722.90) (246.85, 730.56) (234.48, 730.56)      /T1_2 are	<|special_separator|>
-(248.72, 722.90) (284.52, 722.90) (284.52, 730.56) (248.72, 730.56)      /T1_2 averaged	<|special_separator|>
-(286.39, 722.90) (294.75, 722.90) (294.75, 730.56) (286.39, 730.56)      /T1_2 to	<|special_separator|>
-(039.69, 712.15) (137.03, 712.15) (137.03, 719.81) (039.69, 719.81)      /T1_2 obtainthefinalembedding	<|special_separator|>
-(138.88, 710.92) (142.35, 710.92) (142.35, 719.68) (138.88, 719.68)      /C0_3 z	<|special_separator|>
-(142.35, 715.92) (153.91, 715.92) (153.91, 721.65) (142.35, 721.65)      /C0_4 drug	<|special_separator|>
-(142.35, 708.72) (143.89, 708.72) (143.89, 714.85) (142.35, 714.85)      /C0_3 i	<|special_separator|>
-(153.68, 712.15) (204.54, 712.15) (204.54, 719.81) (153.68, 719.81)      /T1_2 .Theresulting	<|special_separator|>
-(206.38, 710.92) (209.85, 710.92) (209.85, 719.68) (206.38, 719.68)      /C0_5 z	<|special_separator|>
-(209.85, 715.92) (221.42, 715.92) (221.42, 721.65) (209.85, 721.65)      /C0_6 drug	<|special_separator|>
-(209.85, 708.72) (211.39, 708.72) (211.39, 714.85) (209.85, 714.85)      /C0_5 i	<|special_separator|>
-(222.14, 712.15) (294.80, 712.15) (294.80, 719.81) (222.14, 719.81)      /T1_2 hasthesamedimen-	<|special_separator|>
-(039.69, 701.40) (174.04, 701.40) (174.04, 709.06) (039.69, 709.06)      /T1_2 sionsasifonedrugwereembedded.	<|special_separator|>
-(039.69, 679.94) (083.69, 679.94) (083.69, 687.92) (039.69, 687.92)      /T1_4 Evaluation	<|special_separator|>
-(039.69, 669.15) (090.44, 669.15) (090.44, 676.81) (039.69, 676.81)      /T1_2 Weevaluated	<|special_separator|>
-(092.08, 669.15) (104.67, 669.15) (104.67, 676.81) (092.08, 676.81)      /T1_2 the	<|special_separator|>
-(106.31, 669.15) (244.12, 669.15) (244.12, 676.81) (106.31, 676.81)      /T1_2 MongeGapandCMongeindifferent	<|special_separator|>
-(245.76, 669.15) (278.45, 669.15) (278.45, 676.81) (245.76, 676.81)      /T1_2 settings.	<|special_separator|>
-(280.09, 669.15) (294.69, 669.15) (294.69, 676.81) (280.09, 676.81)      /T1_2 The	<|special_separator|>
-(039.69, 658.40) (059.40, 658.40) (059.40, 666.06) (039.69, 666.06)      /T1_2 basic	<|special_separator|>
-(061.05, 658.40) (165.47, 658.40) (165.47, 666.06) (061.05, 666.06)      /T1_2 Mongemodelsareunaware	<|special_separator|>
-(167.11, 658.40) (175.11, 658.40) (175.11, 666.06) (167.11, 666.06)      /T1_2 of	<|special_separator|>
-(176.76, 658.40) (220.81, 658.40) (220.81, 666.06) (176.76, 666.06)      /T1_2 conditional	<|special_separator|>
-(222.45, 658.40) (257.89, 658.40) (257.89, 666.06) (222.45, 666.06)      /T1_2 contexts,	<|special_separator|>
-(259.54, 658.40) (273.93, 658.40) (273.93, 666.06) (259.54, 666.06)      /T1_2 and	<|special_separator|>
-(275.58, 658.40) (294.81, 658.40) (294.81, 666.06) (275.58, 666.06)      /T1_2 thus,	<|special_separator|>
-(039.69, 647.65) (294.82, 647.65) (294.82, 655.31) (039.69, 655.31)      /T1_2 thereisasinglemodelpercondition.WhenaMongemodelistrained	<|special_separator|>
-(039.69, 636.90) (294.66, 636.90) (294.66, 644.56) (039.69, 644.56)      /T1_2 onseveralconditions,itdoesnotreceiveanyconditionalinformation	<|special_separator|>
-(039.69, 626.15) (204.64, 626.15) (204.64, 633.81) (039.69, 633.81)      /T1_2 and,therefore,treatsallconditionsasequal.	<|special_separator|>
-(056.69, 615.40) (294.71, 615.40) (294.71, 623.06) (056.69, 623.06)      /T1_2 AllISmodelsweretrainedondataforallconditions,usingan80/20	<|special_separator|>
-(039.69, 604.65) (100.47, 604.65) (100.47, 612.31) (039.69, 612.31)      /T1_2 training/testing	<|special_separator|>
-(102.13, 604.65) (119.07, 604.65) (119.07, 612.31) (102.13, 612.31)      /T1_2 split	<|special_separator|>
-(120.73, 604.65) (132.26, 604.65) (132.26, 612.31) (120.73, 612.31)      /T1_2 for	<|special_separator|>
-(133.93, 604.65) (151.96, 604.65) (151.96, 612.31) (133.93, 612.31)      /T1_2 each	<|special_separator|>
-(153.62, 604.65) (193.35, 604.65) (193.35, 612.31) (153.62, 612.31)      /T1_2 condition.	<|special_separator|>
-(195.02, 604.65) (209.66, 604.65) (209.66, 612.31) (195.02, 612.31)      /T1_2 The	<|special_separator|>
-(211.32, 604.65) (251.74, 604.65) (251.74, 612.31) (211.32, 612.31)      /T1_2 evaluation	<|special_separator|>
-(253.41, 604.65) (294.66, 604.65) (294.66, 612.31) (253.41, 612.31)      /T1_2 conditions	<|special_separator|>
-(039.69, 593.90) (294.77, 593.90) (294.77, 601.56) (039.69, 601.56)      /T1_2 wereseenduringtraining.ForallMongemodels,thismeansthatthe	<|special_separator|>
-(039.69, 583.15) (057.98, 583.15) (057.98, 590.81) (039.69, 590.81)      /T1_2 dose	<|special_separator|>
-(059.64, 583.15) (074.07, 583.15) (074.07, 590.81) (059.64, 590.81)      /T1_2 and	<|special_separator|>
-(075.73, 583.15) (094.19, 583.15) (094.19, 590.81) (075.73, 590.81)      /T1_2 drug	<|special_separator|>
-(095.86, 583.15) (122.69, 583.15) (122.69, 590.81) (095.86, 590.81)      /T1_2 cannot	<|special_separator|>
-(124.35, 583.15) (133.84, 583.15) (133.84, 590.81) (124.35, 590.81)      /T1_2 be	<|special_separator|>
-(135.51, 583.15) (190.29, 583.15) (190.29, 590.81) (135.51, 590.81)      /T1_2 distinguished.	<|special_separator|>
-(191.96, 583.15) (236.69, 583.15) (236.69, 590.81) (191.96, 590.81)      /T1_2 Conversely,	<|special_separator|>
-(238.35, 583.15) (249.88, 583.15) (249.88, 590.81) (238.35, 590.81)      /T1_2 for	<|special_separator|>
-(251.55, 583.15) (294.68, 583.15) (294.68, 590.81) (251.55, 590.81)      /T1_2 allCMonge	<|special_separator|>
-(039.69, 572.40) (070.08, 572.40) (070.08, 580.06) (039.69, 580.06)      /T1_2 models,	<|special_separator|>
-(071.70, 572.40) (084.23, 572.40) (084.23, 580.06) (071.70, 580.06)      /T1_2 the	<|special_separator|>
-(085.85, 572.40) (103.98, 572.40) (103.98, 580.06) (085.85, 580.06)      /T1_2 dose	<|special_separator|>
-(105.60, 572.40) (122.20, 572.40) (122.20, 580.06) (105.60, 580.06)      /T1_2 and,	<|special_separator|>
-(123.82, 572.40) (187.84, 572.40) (187.84, 580.06) (123.82, 580.06)      /T1_2 whenapplicable,	<|special_separator|>
-(189.46, 572.40) (201.98, 572.40) (201.98, 580.06) (189.46, 580.06)      /T1_2 the	<|special_separator|>
-(203.60, 572.40) (224.19, 572.40) (224.19, 580.06) (203.60, 580.06)      /T1_2 drug,	<|special_separator|>
-(225.81, 572.40) (244.08, 572.40) (244.08, 580.06) (225.81, 580.06)      /T1_2 were	<|special_separator|>
-(245.70, 572.40) (278.85, 572.40) (278.85, 580.06) (245.70, 580.06)      /T1_2 encoded	<|special_separator|>
-(280.47, 572.40) (294.79, 572.40) (294.79, 580.06) (280.47, 580.06)      /T1_2 and	<|special_separator|>
-(039.69, 561.65) (192.84, 561.65) (192.84, 569.31) (039.69, 569.31)      /T1_2 thisinformationwasgiventothemodels.	<|special_separator|>
-(056.69, 550.90) (294.80, 550.90) (294.80, 558.56) (056.69, 558.56)      /T1_2 IntheOOSsetting,theMongeandCMongemodelswereevaluated	<|special_separator|>
-(039.69, 540.15) (049.83, 540.15) (049.83, 547.81) (039.69, 547.81)      /T1_2 on	<|special_separator|>
-(051.57, 540.15) (084.96, 540.15) (084.96, 547.81) (051.57, 547.81)      /T1_2 held-out	<|special_separator|>
-(086.69, 540.15) (128.62, 540.15) (128.62, 547.81) (086.69, 547.81)      /T1_2 conditions	<|special_separator|>
-(130.36, 540.15) (144.99, 540.15) (144.99, 547.81) (130.36, 547.81)      /T1_2 and	<|special_separator|>
-(146.73, 540.15) (175.08, 540.15) (175.08, 547.81) (146.73, 547.81)      /T1_2 trained	<|special_separator|>
-(176.82, 540.15) (186.96, 540.15) (186.96, 547.81) (176.82, 547.81)      /T1_2 on	<|special_separator|>
-(188.70, 540.15) (197.94, 540.15) (197.94, 547.81) (188.70, 547.81)      /T1_2 all	<|special_separator|>
-(199.68, 540.15) (221.17, 540.15) (221.17, 547.81) (199.68, 547.81)      /T1_2 other	<|special_separator|>
-(222.91, 540.15) (267.16, 540.15) (267.16, 547.81) (222.91, 547.81)      /T1_2 conditions.	<|special_separator|>
-(268.90, 540.15) (294.63, 540.15) (294.63, 547.81) (268.90, 547.81)      /T1_2 Unless	<|special_separator|>
-(039.69, 529.40) (077.60, 529.40) (077.60, 537.06) (039.69, 537.06)      /T1_2 specified,	<|special_separator|>
-(079.31, 529.40) (130.40, 529.40) (130.40, 537.06) (079.31, 537.06)      /T1_2 weemployed	<|special_separator|>
-(132.12, 529.40) (136.45, 529.40) (136.45, 537.06) (132.12, 537.06)      /T1_2 a	<|special_separator|>
-(138.16, 529.40) (191.57, 529.40) (191.57, 537.06) (138.16, 537.06)      /T1_2 leave-one-out	<|special_separator|>
-(193.28, 529.40) (220.39, 529.40) (220.39, 537.06) (193.28, 537.06)      /T1_2 setting	<|special_separator|>
-(222.10, 529.40) (245.83, 529.40) (245.83, 537.06) (222.10, 537.06)      /T1_2 where	<|special_separator|>
-(247.55, 529.40) (288.16, 529.40) (288.16, 537.06) (247.55, 537.06)      /T1_2 wetrained	<|special_separator|>
-(289.88, 529.41) (294.85, 529.41) (294.85, 537.06) (289.88, 537.06)      /T1_3 n	<|special_separator|>
-(039.69, 518.65) (067.89, 518.65) (067.89, 526.31) (039.69, 526.31)      /T1_2 models	<|special_separator|>
-(069.52, 518.65) (081.00, 518.65) (081.00, 526.31) (069.52, 526.31)      /T1_2 for	<|special_separator|>
-(082.63, 518.66) (087.56, 518.66) (087.56, 526.31) (082.63, 526.31)      /T1_3 n	<|special_separator|>
-(089.17, 518.65) (121.84, 518.65) (121.84, 526.31) (089.17, 526.31)      /T1_2 held-out	<|special_separator|>
-(123.47, 518.65) (166.79, 518.65) (166.79, 526.31) (123.47, 526.31)      /T1_2 conditions,	<|special_separator|>
-(168.42, 518.65) (193.65, 518.65) (193.65, 526.31) (168.42, 526.31)      /T1_2 always	<|special_separator|>
-(195.28, 518.65) (222.89, 518.65) (222.89, 526.31) (195.28, 526.31)      /T1_2 leaving	<|special_separator|>
-(224.52, 518.65) (238.92, 518.65) (238.92, 526.31) (224.52, 526.31)      /T1_2 one	<|special_separator|>
-(240.55, 518.65) (277.80, 518.65) (277.80, 526.31) (240.55, 526.31)      /T1_2 condition	<|special_separator|>
-(279.43, 518.65) (294.87, 518.65) (294.87, 526.31) (279.43, 526.31)      /T1_2 out.	<|special_separator|>
-(039.69, 507.90) (294.81, 507.90) (294.81, 515.56) (039.69, 515.56)      /T1_2 AsintheISsetting,theMongemodelsdonotdistinguishconditionsin	<|special_separator|>
-(039.69, 497.15) (294.88, 497.15) (294.88, 504.81) (039.69, 504.81)      /T1_2 thetrainingorevaluationsetting,whereasCMongecanbeconditioned	<|special_separator|>
-(039.69, 486.40) (157.24, 486.40) (157.24, 494.06) (039.69, 494.06)      /T1_2 ondosealoneordruganddose.	<|special_separator|>
-(056.69, 475.65) (294.62, 475.65) (294.62, 483.31) (056.69, 483.31)      /T1_2 First,wetestedtheabilityofamodeltoconditiononthedoseand	<|special_separator|>
-(039.69, 464.90) (047.88, 464.90) (047.88, 472.56) (039.69, 472.56)      /T1_2 to	<|special_separator|>
-(049.53, 464.90) (089.17, 464.90) (089.17, 472.56) (049.53, 472.56)      /T1_2 generalize	<|special_separator|>
-(090.83, 464.90) (099.02, 464.90) (099.02, 472.56) (090.83, 472.56)      /T1_2 to	<|special_separator|>
-(100.68, 464.90) (128.22, 464.90) (128.22, 472.56) (100.68, 472.56)      /T1_2 unseen	<|special_separator|>
-(129.87, 464.90) (154.20, 464.90) (154.20, 472.56) (129.87, 472.56)      /T1_2 doses.	<|special_separator|>
-(155.86, 464.90) (294.74, 464.90) (294.74, 472.56) (155.86, 472.56)      /T1_2 Therefore,theCMongemodelswere	<|special_separator|>
-(039.69, 454.15) (294.82, 454.15) (294.82, 461.81) (039.69, 461.81)      /T1_2 conditionedondose(Dose)inbothanISsettingandanOOSsetting.	<|special_separator|>
-(039.69, 443.40) (061.95, 443.40) (061.95, 451.06) (039.69, 451.06)      /T1_2 Then,	<|special_separator|>
-(063.68, 443.40) (100.55, 443.40) (100.55, 451.06) (063.68, 451.06)      /T1_2 wetested	<|special_separator|>
-(102.28, 443.40) (115.09, 443.40) (115.09, 451.06) (102.28, 451.06)      /T1_2 the	<|special_separator|>
-(116.82, 443.40) (141.42, 443.40) (141.42, 451.06) (116.82, 451.06)      /T1_2 ability	<|special_separator|>
-(143.15, 443.40) (151.27, 443.40) (151.27, 451.06) (143.15, 451.06)      /T1_2 of	<|special_separator|>
-(153.01, 443.40) (165.81, 443.40) (165.81, 451.06) (153.01, 451.06)      /T1_2 the	<|special_separator|>
-(167.55, 443.40) (192.45, 443.40) (192.45, 451.06) (167.55, 451.06)      /T1_2 model	<|special_separator|>
-(194.19, 443.40) (202.49, 443.40) (202.49, 451.06) (194.19, 451.06)      /T1_2 to	<|special_separator|>
-(204.23, 443.40) (242.27, 443.40) (242.27, 451.06) (204.23, 451.06)      /T1_2 condition	<|special_separator|>
-(244.01, 443.40) (254.14, 443.40) (254.14, 451.06) (244.01, 451.06)      /T1_2 on	<|special_separator|>
-(255.88, 443.40) (278.41, 443.40) (278.41, 451.06) (255.88, 451.06)      /T1_2 drugs	<|special_separator|>
-(280.15, 443.40) (294.77, 443.40) (294.77, 451.06) (280.15, 451.06)      /T1_2 and	<|special_separator|>
-(039.69, 432.65) (294.77, 432.65) (294.77, 440.31) (039.69, 440.31)      /T1_2 generalizetounseendrugsbyholdingoutalldosesofonedrugduring	<|special_separator|>
-(039.69, 421.90) (294.77, 421.90) (294.77, 429.56) (039.69, 429.56)      /T1_2 training.TheCMongemodelswereconditionedonbothdruganddose	<|special_separator|>
-(039.69, 411.15) (083.81, 411.15) (083.81, 418.81) (039.69, 418.81)      /T1_2 (DrugDose)	<|special_separator|>
-(085.32, 411.15) (092.61, 411.15) (092.61, 418.81) (085.32, 418.81)      /T1_2 in	<|special_separator|>
-(094.12, 411.15) (111.92, 411.15) (111.92, 418.81) (094.12, 418.81)      /T1_2 anIS	<|special_separator|>
-(113.43, 411.15) (139.88, 411.15) (139.88, 418.81) (113.43, 418.81)      /T1_2 setting	<|special_separator|>
-(141.39, 411.15) (234.71, 411.15) (234.71, 418.81) (141.39, 418.81)      /T1_2 andanOOSsetting.Last,	<|special_separator|>
-(236.22, 411.15) (294.79, 411.15) (294.79, 418.81) (236.22, 418.81)      /T1_2 weinvestigated	<|special_separator|>
-(039.69, 400.40) (294.75, 400.40) (294.75, 408.06) (039.69, 408.06)      /T1_2 modelperformanceonconditioningonDrugDosewithmoretraining	<|special_separator|>
-(039.69, 389.65) (294.84, 389.65) (294.84, 397.31) (039.69, 397.31)      /T1_2 drugs,bothinanISsettingandanOOSsetting.IntheOOSsetting,we	<|special_separator|>
-(039.69, 378.90) (294.75, 378.90) (294.75, 386.56) (039.69, 386.56)      /T1_2 increasedtheOOSconditionsbyleavingninedrugsoutinsteadofone.	<|special_separator|>
-(056.69, 368.15) (119.77, 368.15) (119.77, 375.81) (056.69, 375.81)      /T1_2 Weevaluatedthe	<|special_separator|>
-(120.84, 368.16) (125.98, 368.16) (125.98, 375.81) (120.84, 375.81)      /T1_3 R	<|special_separator|>
-(126.01, 372.34) (128.69, 372.34) (128.69, 376.93) (126.01, 376.93)      /T1_2 2	<|special_separator|>
-(129.76, 368.15) (294.70, 368.15) (294.70, 375.81) (129.76, 375.81)      /T1_2 metricbetweentheperturbedandpredicted	<|special_separator|>
-(039.69, 357.40) (294.81, 357.40) (294.81, 365.06) (039.69, 365.06)      /T1_2 featuremeans,theWassersteindistance(equation(2))andtheMMD.	<|special_separator|>
-(039.69, 346.65) (294.78, 346.65) (294.78, 354.31) (039.69, 354.31)      /T1_2 Allmetricswerecalculatedonbatchesofobservationssampledfrom	<|special_separator|>
-(039.69, 335.90) (294.80, 335.90) (294.80, 343.56) (039.69, 343.56)      /T1_2 thetestsetandwereportthemeanacrossthesebatches.Specifically,	<|special_separator|>
-(039.69, 325.15) (294.86, 325.15) (294.86, 332.81) (039.69, 332.81)      /T1_2 foreachcondition,wealwaysperformednineindependentevaluations	<|special_separator|>
-(039.69, 314.40) (294.66, 314.40) (294.66, 322.06) (039.69, 322.06)      /T1_2 byrandomlysamplingasetofcontrolcells.Thus,allreportedmetrics	<|special_separator|>
-(039.69, 303.65) (042.75, 303.65) (042.75, 311.31) (039.69, 311.31)      /T1_2 (	<|special_separator|>
-(042.78, 303.66) (047.98, 303.66) (047.98, 311.31) (042.78, 311.31)      /T1_3 R	<|special_separator|>
-(048.01, 307.84) (050.71, 307.84) (050.71, 312.43) (048.01, 312.43)      /T1_2 2	<|special_separator|>
-(050.73, 303.65) (053.02, 303.65) (053.02, 311.31) (050.73, 311.31)      /T1_2 ,	<|special_separator|>
-(054.69, 303.65) (139.56, 303.65) (139.56, 311.31) (054.69, 311.31)      /T1_2 MMDandWasserstein	<|special_separator|>
-(141.23, 303.65) (176.91, 303.65) (176.91, 311.31) (141.23, 311.31)      /T1_2 distance)	<|special_separator|>
-(178.58, 303.65) (197.04, 303.65) (197.04, 311.31) (178.58, 311.31)      /T1_2 were	<|special_separator|>
-(198.71, 303.65) (238.51, 303.65) (238.51, 311.31) (198.71, 311.31)      /T1_2 computed	<|special_separator|>
-(240.19, 303.65) (253.28, 303.65) (253.28, 311.31) (240.19, 311.31)      /T1_2 per	<|special_separator|>
-(254.95, 303.65) (294.84, 303.65) (294.84, 311.31) (254.95, 311.31)      /T1_2 condition,	<|special_separator|>
-(039.69, 292.90) (294.82, 292.90) (294.82, 300.56) (039.69, 300.56)      /T1_2 includingthisbootstrappingprocedure,andthenaveragedacrossall	<|special_separator|>
-(039.69, 282.15) (294.87, 282.15) (294.87, 289.81) (039.69, 289.81)      /T1_2 conditions(ifany).Theexperimentsusedthesamehyperparameters,	<|special_separator|>
-(039.69, 271.40) (294.78, 271.40) (294.78, 279.06) (039.69, 279.06)      /T1_2 exceptforthenumberofoptimizationstepsandthelatentencoding	<|special_separator|>
-(039.69, 260.65) (047.68, 260.65) (047.68, 268.31) (039.69, 268.31)      /T1_2 of	<|special_separator|>
-(049.32, 260.65) (061.90, 260.65) (061.90, 268.31) (049.32, 268.31)      /T1_2 the	<|special_separator|>
-(063.55, 260.65) (094.84, 260.65) (094.84, 268.31) (063.55, 268.31)      /T1_2 features	<|special_separator|>
-(096.49, 260.65) (110.87, 260.65) (110.87, 268.31) (096.49, 268.31)      /T1_2 and	<|special_separator|>
-(112.51, 260.65) (141.85, 260.65) (141.85, 268.31) (112.51, 268.31)      /T1_2 context	<|special_separator|>
-(143.50, 260.65) (158.05, 260.65) (158.05, 268.31) (143.50, 268.31)      /T1_2 (for	<|special_separator|>
-(159.70, 260.65) (209.53, 260.65) (209.53, 268.31) (159.70, 268.31)      /T1_2 moredetails,	<|special_separator|>
-(211.17, 260.65) (223.71, 260.65) (223.71, 268.31) (211.17, 268.31)      /T1_2 see	<|special_separator|>
-(225.35, 260.65) (294.67, 260.65) (294.67, 268.31) (225.35, 268.31)      /T1_2 'Hyperparameters	<|special_separator|>
-(039.69, 249.90) (294.78, 249.90) (294.78, 257.56) (039.69, 257.56)      /T1_2 andmodelsizes'inMethods).FortheSciPlexdataset,wetrainedand	<|special_separator|>
-(039.69, 239.15) (203.35, 239.15) (203.35, 246.81) (039.69, 246.81)      /T1_2 evaluatedtheCMinthefollowingscenarios:	<|special_separator|>
-(039.78, 224.40) (049.22, 224.40) (049.22, 232.06) (039.78, 232.06)      /T1_2 (1)	<|special_separator|>
-(056.13, 224.40) (290.12, 224.40) (290.12, 232.06) (056.13, 232.06)      /T1_2 Monge:Asahypotheticalupperboundonperformance,wefit-	<|special_separator|>
-(056.13, 213.65) (293.84, 213.65) (293.84, 221.31) (056.13, 221.31)      /T1_2 tedseparateMongeGapmodels,oneperdrug-dosepair.These	<|special_separator|>
-(056.13, 202.90) (251.43, 202.90) (251.43, 210.56) (056.13, 210.56)      /T1_2 modelsdonothaveanycontext(UsciddaandCuturi	<|special_separator|>
-(251.43, 207.09) (256.74, 207.09) (256.74, 211.68) (251.43, 211.68)      /T1_2 32	<|special_separator|>
-(256.74, 202.90) (262.01, 202.90) (262.01, 210.56) (256.74, 210.56)      /T1_2 ).	<|special_separator|>
-(039.78, 192.15) (050.24, 192.15) (050.24, 199.81) (039.78, 199.81)      /T1_2 (2)	<|special_separator|>
-(056.13, 192.15) (280.72, 192.15) (280.72, 199.81) (056.13, 199.81)      /T1_2 Monge-DrugandMonge-DrugDose:Tomotivatethecontex-	<|special_separator|>
-(056.13, 181.40) (287.60, 181.40) (287.60, 189.06) (056.13, 189.06)      /T1_2 tualsettings,wefittedaMongeGapmodelthatwastrainedon	<|special_separator|>
-(056.13, 170.65) (082.66, 170.65) (082.66, 178.31) (056.13, 178.31)      /T1_2 several	<|special_separator|>
-(084.11, 170.65) (124.82, 170.65) (124.82, 178.31) (084.11, 178.31)      /T1_2 conditions	<|special_separator|>
-(126.27, 170.65) (139.38, 170.65) (139.38, 178.31) (126.27, 178.31)      /T1_2 but	<|special_separator|>
-(140.83, 170.65) (292.68, 170.65) (292.68, 178.31) (140.83, 178.31)      /T1_2 wasunawareofconditionalinformation,	<|special_separator|>
-(056.13, 159.90) (260.91, 159.90) (260.91, 167.56) (056.13, 167.56)      /T1_2 andweevaluateditonconditionsseenduringtraining.	<|special_separator|>
-(053.58, 146.31) (056.60, 146.31) (056.60, 153.97) (053.58, 153.97)      /T1_2 •	<|special_separator|>
-(067.47, 146.31) (290.77, 146.31) (290.77, 153.97) (067.47, 153.97)      /T1_2 Monge-Dose-IS:ahomogeneousmodelforeachdrug,using	<|special_separator|>
-(067.47, 135.56) (084.25, 135.56) (084.25, 143.22) (067.47, 143.22)      /T1_2 data	<|special_separator|>
-(085.70, 135.56) (114.87, 135.56) (114.87, 143.22) (085.70, 143.22)      /T1_2 fromall	<|special_separator|>
-(116.32, 135.56) (132.65, 135.56) (132.65, 143.22) (116.32, 143.22)      /T1_2 four	<|special_separator|>
-(134.11, 135.56) (158.14, 135.56) (158.14, 143.22) (134.11, 143.22)      /T1_2 doses.	<|special_separator|>
-(053.58, 124.81) (056.60, 124.81) (056.60, 132.47) (053.58, 132.47)      /T1_2 •	<|special_separator|>
-(067.47, 124.81) (289.59, 124.81) (289.59, 132.47) (067.47, 132.47)      /T1_2 Monge-DrugDose-IS:onemodelonallconditions(alldrug-	<|special_separator|>
-(067.47, 114.06) (111.06, 114.06) (111.06, 121.72) (067.47, 121.72)      /T1_2 dosepairs).	<|special_separator|>
-(053.58, 103.31) (056.60, 103.31) (056.60, 110.97) (053.58, 110.97)      /T1_2 •	<|special_separator|>
-(067.47, 103.31) (275.26, 103.31) (275.26, 110.97) (067.47, 110.97)      /T1_2 Monge-Dose-OOS:amodelforeachdrug,withdifferent	<|special_separator|>
-(067.47, 092.56) (089.28, 092.56) (089.28, 100.22) (067.47, 100.22)      /T1_2 doses	<|special_separator|>
-(090.73, 092.56) (103.56, 092.56) (103.56, 100.22) (090.73, 100.22)      /T1_2 left	<|special_separator|>
-(105.02, 092.56) (118.11, 092.56) (118.11, 100.22) (105.02, 100.22)      /T1_2 out	<|special_separator|>
-(119.56, 092.56) (145.09, 092.56) (145.09, 100.22) (119.56, 100.22)      /T1_2 during	<|special_separator|>
-(146.54, 092.56) (178.91, 092.56) (178.91, 100.22) (146.54, 100.22)      /T1_2 training.	<|special_separator|>
-(053.58, 081.82) (056.60, 081.82) (056.60, 089.47) (053.58, 089.47)      /T1_2 •	<|special_separator|>
-(067.47, 081.82) (280.80, 081.82) (280.80, 089.47) (067.47, 089.47)      /T1_2 Monge-DrugDose-OOS:amodeltrainedonallconditions	<|special_separator|>
-(067.47, 071.07) (080.58, 071.07) (080.58, 078.72) (067.47, 078.72)      /T1_2 but	<|special_separator|>
-(082.03, 071.07) (094.51, 071.07) (094.51, 078.72) (082.03, 078.72)      /T1_2 the	<|special_separator|>
-(095.97, 071.07) (117.77, 071.07) (117.77, 078.72) (095.97, 078.72)      /T1_2 doses	<|special_separator|>
-(119.22, 071.07) (127.15, 071.07) (127.15, 078.72) (119.22, 078.72)      /T1_2 of	<|special_separator|>
-(128.60, 071.07) (141.09, 071.07) (141.09, 078.72) (128.60, 078.72)      /T1_2 the	<|special_separator|>
-(142.54, 071.07) (174.96, 071.07) (174.96, 078.72) (142.54, 078.72)      /T1_2 held-out	<|special_separator|>
-(176.41, 071.07) (206.69, 071.07) (206.69, 078.72) (176.41, 078.72)      /T1_2 drug(s).	<|special_separator|>
-(039.79, 057.31) (050.05, 057.31) (050.05, 064.97) (039.79, 064.97)      /T1_2 (3)	<|special_separator|>
-(056.14, 057.31) (287.55, 057.31) (287.55, 064.97) (056.14, 064.97)      /T1_2 CMonge-Dose-ISandCMonge-Dose-OOS:Wefittedconditional	<|special_separator|>
-(056.14, 046.56) (294.70, 046.56) (294.70, 054.22) (056.14, 054.22)      /T1_2 modelsforeachdrugwiththescalardoseascontext.TheISsetting	<|special_separator|>
-(322.59, 733.67) (556.38, 733.67) (556.38, 741.33) (322.59, 741.33)      /T1_2 sawalldosesduringtraining.FortheOOSsetting,weleftoutdiffer-	<|special_separator|>
-(322.59, 722.92) (561.16, 722.92) (561.16, 730.58) (322.59, 730.58)      /T1_2 entdosesduringtraining,thuscreatinginterpolationandextrapola-	<|special_separator|>
-(322.59, 712.17) (368.64, 712.17) (368.64, 719.83) (322.59, 719.83)      /T1_2 tionsettings.	<|special_separator|>
-(306.25, 701.42) (316.94, 701.42) (316.94, 709.08) (306.25, 709.08)      /T1_2 (4)	<|special_separator|>
-(322.59, 701.42) (554.28, 701.42) (554.28, 709.08) (322.59, 709.08)      /T1_2 CMonge-DrugDose-RDKitandCMonge-DrugDose-MoA:Asin-	<|special_separator|>
-(322.59, 690.67) (552.17, 690.67) (552.17, 698.33) (322.59, 698.33)      /T1_2 glemodelwasfittedtoalldata,conditionedondruganddose	<|special_separator|>
-(322.59, 679.92) (542.51, 679.92) (542.51, 687.58) (322.59, 687.58)      /T1_2 context.Toencodethedrug,wecomparedthefingerprints	<|special_separator|>
-(322.59, 669.17) (350.86, 669.17) (350.86, 676.83) (322.59, 676.83)      /T1_2 (RDKit)	<|special_separator|>
-(352.31, 669.17) (360.42, 669.17) (360.42, 676.83) (352.31, 676.83)      /T1_2 to	<|special_separator|>
-(361.87, 669.17) (366.15, 669.17) (366.15, 676.83) (361.87, 676.83)      /T1_2 a	<|special_separator|>
-(367.60, 669.17) (476.46, 669.17) (476.46, 676.83) (367.60, 676.83)      /T1_2 data-drivenapproach(MoA).	<|special_separator|>
-(320.04, 651.59) (323.06, 651.59) (323.06, 659.25) (320.04, 659.25)      /T1_2 •	<|special_separator|>
-(333.94, 651.59) (425.31, 651.59) (425.31, 659.25) (333.94, 659.25)      /T1_2 CMonge-DrugDose-x-IS:	<|special_separator|>
-(426.77, 651.59) (544.53, 651.59) (544.53, 659.25) (426.77, 659.25)      /T1_2 Allconditionswereseenduring	<|special_separator|>
-(333.94, 640.84) (366.31, 640.84) (366.31, 648.50) (333.94, 648.50)      /T1_2 training.	<|special_separator|>
-(320.04, 630.09) (323.06, 630.09) (323.06, 637.75) (320.04, 637.75)      /T1_2 •	<|special_separator|>
-(333.94, 630.09) (550.31, 630.09) (550.31, 637.75) (333.94, 637.75)      /T1_2 CMonge-DrugDose-x-OOS:Alldosesofoneormoredrugs	<|special_separator|>
-(333.94, 619.34) (428.91, 619.34) (428.91, 627.00) (333.94, 627.00)      /T1_2 wereheldduringtraining	<|special_separator|>
-(430.36, 619.34) (441.76, 619.34) (441.76, 627.00) (430.36, 627.00)      /T1_2 for	<|special_separator|>
-(443.21, 619.34) (485.32, 619.34) (485.32, 627.00) (443.21, 627.00)      /T1_2 evaluation.	<|special_separator|>
-(323.16, 604.66) (561.28, 604.66) (561.28, 612.32) (323.16, 612.32)      /T1_2 FortheconditionalMonge4iexperiments,combinatorialthera-	<|special_separator|>
-(306.14, 593.90) (321.84, 593.90) (321.84, 601.56) (306.14, 601.56)      /T1_2 pies	<|special_separator|>
-(323.48, 593.90) (367.49, 593.90) (367.49, 601.56) (323.48, 601.56)      /T1_2 (conditions	<|special_separator|>
-(369.12, 593.90) (385.90, 593.90) (385.90, 601.56) (369.12, 601.56)      /T1_2 with	<|special_separator|>
-(387.53, 593.90) (401.90, 593.90) (401.90, 601.56) (387.53, 601.56)      /T1_2 two	<|special_separator|>
-(403.53, 593.90) (412.00, 593.90) (412.00, 601.56) (403.53, 601.56)      /T1_2 or	<|special_separator|>
-(413.63, 593.90) (433.96, 593.90) (433.96, 601.56) (413.63, 601.56)      /T1_2 three	<|special_separator|>
-(435.59, 593.90) (460.75, 593.90) (460.75, 601.56) (435.59, 601.56)      /T1_2 drugs)	<|special_separator|>
-(462.38, 593.90) (480.71, 593.90) (480.71, 601.56) (462.38, 601.56)      /T1_2 were	<|special_separator|>
-(482.34, 593.90) (513.60, 593.90) (513.60, 601.56) (482.34, 601.56)      /T1_2 handled	<|special_separator|>
-(515.23, 593.90) (548.17, 593.90) (548.17, 601.56) (515.23, 601.56)      /T1_2 similarly	<|special_separator|>
-(549.80, 593.90) (561.28, 593.90) (561.28, 601.56) (549.80, 601.56)      /T1_2 for	<|special_separator|>
-(306.14, 583.15) (318.67, 583.15) (318.67, 590.81) (306.14, 590.81)      /T1_2 the	<|special_separator|>
-(320.28, 583.15) (342.56, 583.15) (342.56, 590.81) (320.28, 590.81)      /T1_2 RDKit	<|special_separator|>
-(344.18, 583.15) (358.50, 583.15) (358.50, 590.81) (344.18, 590.81)      /T1_2 and	<|special_separator|>
-(360.11, 583.15) (372.64, 583.15) (372.64, 590.81) (360.11, 590.81)      /T1_2 the	<|special_separator|>
-(374.26, 583.15) (506.11, 583.15) (506.11, 590.81) (374.26, 590.81)      /T1_2 MoAembeddings.Foreachdrugin	<|special_separator|>
-(507.73, 583.15) (520.25, 583.15) (520.25, 590.81) (507.73, 590.81)      /T1_2 the	<|special_separator|>
-(521.87, 583.15) (561.27, 583.15) (561.27, 590.81) (521.87, 590.81)      /T1_2 condition,	<|special_separator|>
-(306.14, 572.40) (318.62, 572.40) (318.62, 580.06) (306.14, 580.06)      /T1_2 the	<|special_separator|>
-(320.22, 572.40) (561.29, 572.40) (561.29, 580.06) (320.22, 580.06)      /T1_2 embeddingwascomputedfortheMoAbasedonthesingle-drug	<|special_separator|>
-(306.14, 561.65) (346.85, 561.65) (346.85, 569.31) (306.14, 569.31)      /T1_2 condition,	<|special_separator|>
-(348.64, 561.65) (354.16, 561.65) (354.16, 569.31) (348.64, 569.31)      /T1_2 if	<|special_separator|>
-(355.96, 561.65) (391.02, 561.65) (391.02, 569.31) (355.96, 569.31)      /T1_2 possible.	<|special_separator|>
-(392.82, 561.65) (415.22, 561.65) (415.22, 569.31) (392.82, 569.31)      /T1_2 Then,	<|special_separator|>
-(417.02, 561.65) (429.90, 561.65) (429.90, 569.31) (417.02, 569.31)      /T1_2 the	<|special_separator|>
-(431.70, 561.65) (480.51, 561.65) (480.51, 569.31) (431.70, 569.31)      /T1_2 embeddings	<|special_separator|>
-(482.30, 561.65) (501.64, 561.65) (501.64, 569.31) (482.30, 569.31)      /T1_2 from	<|special_separator|>
-(503.44, 561.65) (512.75, 561.65) (512.75, 569.31) (503.44, 569.31)      /T1_2 all	<|special_separator|>
-(514.54, 561.65) (537.22, 561.65) (537.22, 569.31) (514.54, 569.31)      /T1_2 drugs	<|special_separator|>
-(539.02, 561.65) (546.52, 561.65) (546.52, 569.31) (539.02, 569.31)      /T1_2 in	<|special_separator|>
-(548.32, 561.65) (561.20, 561.65) (561.20, 569.31) (548.32, 569.31)      /T1_2 the	<|special_separator|>
-(306.14, 550.90) (343.44, 550.90) (343.44, 558.56) (306.14, 558.56)      /T1_2 condition	<|special_separator|>
-(345.08, 550.90) (363.43, 550.90) (363.43, 558.56) (345.08, 558.56)      /T1_2 were	<|special_separator|>
-(365.07, 550.90) (391.36, 550.90) (391.36, 558.56) (365.07, 558.56)      /T1_2 passed	<|special_separator|>
-(393.00, 550.90) (424.33, 550.90) (424.33, 558.56) (393.00, 558.56)      /T1_2 through	<|special_separator|>
-(425.97, 550.90) (438.55, 550.90) (438.55, 558.56) (425.97, 558.56)      /T1_2 the	<|special_separator|>
-(440.19, 550.90) (484.47, 550.90) (484.47, 558.56) (440.19, 558.56)      /T1_2 samedense	<|special_separator|>
-(486.11, 550.90) (506.86, 550.90) (506.86, 558.56) (486.11, 558.56)      /T1_2 layer,	<|special_separator|>
-(508.50, 550.90) (525.30, 550.90) (525.30, 558.56) (508.50, 558.56)      /T1_2 with	<|special_separator|>
-(526.94, 550.90) (561.18, 550.90) (561.18, 558.56) (526.94, 558.56)      /T1_2 thesame	<|special_separator|>
-(306.14, 540.15) (561.40, 540.15) (561.40, 547.81) (306.14, 547.81)      /T1_2 parametersforeachsingle-drugembedding.Weleftoutthecondition	<|special_separator|>
-(306.14, 529.40) (561.27, 529.40) (561.27, 537.06) (306.14, 537.06)      /T1_2 vemurafenib-cobimetinib,asneitherdrugwasmeasuredinisolation	<|special_separator|>
-(306.14, 518.65) (561.23, 518.65) (561.23, 526.31) (306.14, 526.31)      /T1_2 and,therefore,noMoAembeddingwaspossibleforeithersingledrug.	<|special_separator|>
-(306.14, 507.90) (334.86, 507.90) (334.86, 515.56) (306.14, 515.56)      /T1_2 Wealso	<|special_separator|>
-(336.46, 507.90) (349.33, 507.90) (349.33, 515.56) (336.46, 515.56)      /T1_2 left	<|special_separator|>
-(350.94, 507.90) (364.06, 507.90) (364.06, 515.56) (350.94, 515.56)      /T1_2 out	<|special_separator|>
-(365.67, 507.90) (533.41, 507.90) (533.41, 515.56) (365.67, 515.56)      /T1_2 pomalidomide-carfilzomib-dexamethasone,	<|special_separator|>
-(535.02, 507.90) (543.04, 507.90) (543.04, 515.56) (535.02, 515.56)      /T1_2 as	<|special_separator|>
-(544.65, 507.90) (561.29, 507.90) (561.29, 515.56) (544.65, 515.56)      /T1_2 only	<|special_separator|>
-(306.14, 497.15) (492.33, 497.15) (492.33, 504.81) (306.14, 504.81)      /T1_2 dexamethasonewaspresentasasingletreatment.	<|special_separator|>
-(306.14, 475.69) (341.89, 475.69) (341.89, 483.67) (306.14, 483.67)      /T1_4 Datasets	<|special_separator|>
-(306.14, 464.90) (561.10, 464.90) (561.10, 472.56) (306.14, 472.56)      /T1_2 Weusedtwodatasetstotrainandevaluateourmodels.First,theSciPlex	<|special_separator|>
-(306.14, 454.15) (333.38, 454.15) (333.38, 461.81) (306.14, 461.81)      /T1_2 dataset	<|special_separator|>
-(333.35, 458.34) (336.01, 458.34) (336.01, 462.93) (333.35, 462.93)      /T1_2 3	<|special_separator|>
-(337.02, 454.15) (561.20, 454.15) (561.20, 461.81) (337.02, 461.81)      /T1_2 containssingle-cellprofilesfromthreehumancancercelllines	<|special_separator|>
-(306.14, 443.40) (561.26, 443.40) (561.26, 451.06) (306.14, 451.06)      /T1_2 (A549,K562andMCF7)thatwereexposedto188compounds,compris-	<|special_separator|>
-(306.14, 432.65) (318.28, 432.65) (318.28, 440.31) (306.14, 440.31)      /T1_2 ing	<|special_separator|>
-(319.91, 432.65) (332.47, 432.65) (332.47, 440.31) (319.91, 440.31)      /T1_2 187	<|special_separator|>
-(334.10, 432.65) (356.14, 432.65) (356.14, 440.31) (334.10, 440.31)      /T1_2 drugs	<|special_separator|>
-(357.76, 432.65) (373.37, 432.65) (373.37, 440.31) (357.76, 440.31)      /T1_2 that	<|special_separator|>
-(374.99, 432.65) (393.28, 432.65) (393.28, 440.31) (374.99, 440.31)      /T1_2 were	<|special_separator|>
-(394.90, 432.65) (445.55, 432.65) (445.55, 440.31) (394.90, 440.31)      /T1_2 administered	<|special_separator|>
-(447.17, 432.65) (454.64, 432.65) (454.64, 440.31) (447.17, 440.31)      /T1_2 at	<|special_separator|>
-(456.26, 432.65) (472.67, 432.65) (472.67, 440.31) (456.26, 440.31)      /T1_2 four	<|special_separator|>
-(474.30, 432.65) (507.83, 432.65) (507.83, 440.31) (474.30, 440.31)      /T1_2 different	<|special_separator|>
-(509.45, 432.65) (531.37, 432.65) (531.37, 440.31) (509.45, 440.31)      /T1_2 doses	<|special_separator|>
-(532.99, 432.65) (561.12, 432.65) (561.12, 440.31) (532.99, 440.31)      /T1_2 (10nM,	<|special_separator|>
-(306.14, 421.90) (320.86, 421.90) (320.86, 429.56) (306.14, 429.56)      /T1_2 100	<|special_separator|>
-(322.51, 421.90) (337.54, 421.90) (337.54, 429.56) (322.51, 429.56)      /T1_2 nM,	<|special_separator|>
-(339.30, 421.90) (493.11, 421.90) (493.11, 429.56) (339.30, 429.56)      /T1_2 1,000nMand10,000nM)andacontrol	<|special_separator|>
-(494.87, 421.90) (529.73, 421.90) (529.73, 429.56) (494.87, 429.56)      /T1_2 solution.	<|special_separator|>
-(531.49, 421.90) (561.26, 421.90) (561.26, 429.56) (531.49, 429.56)      /T1_2 Weper-	<|special_separator|>
-(306.14, 411.15) (561.14, 411.15) (561.14, 418.81) (306.14, 418.81)      /T1_2 formedexperimentsonaselectionofninedifferentdrugs,asinUscidda	<|special_separator|>
-(306.14, 400.40) (346.31, 400.40) (346.31, 408.06) (306.14, 408.06)      /T1_2 andCuturi	<|special_separator|>
-(346.31, 404.59) (351.63, 404.59) (351.63, 409.18) (346.31, 409.18)      /T1_2 32	<|special_separator|>
-(351.63, 400.40) (353.90, 400.40) (353.90, 408.06) (351.63, 408.06)      /T1_2 ,	<|special_separator|>
-(355.42, 400.40) (409.30, 400.40) (409.30, 408.06) (355.42, 408.06)      /T1_2 andonthefull	<|special_separator|>
-(410.82, 400.40) (432.39, 400.40) (432.39, 408.06) (410.82, 408.06)      /T1_2 range	<|special_separator|>
-(433.92, 400.40) (441.84, 400.40) (441.84, 408.06) (433.92, 408.06)      /T1_2 of	<|special_separator|>
-(443.37, 400.40) (535.41, 400.40) (535.41, 408.06) (443.37, 408.06)      /T1_2 compounds,asinHetzel	<|special_separator|>
-(536.93, 400.40) (544.48, 400.40) (544.48, 408.06) (536.93, 408.06)      /T1_2 et	<|special_separator|>
-(546.01, 400.40) (554.86, 400.40) (554.86, 408.06) (546.01, 408.06)      /T1_2 al.	<|special_separator|>
-(554.89, 404.59) (559.03, 404.59) (559.03, 409.18) (554.89, 409.18)      /T1_2 11	<|special_separator|>
-(559.03, 400.40) (561.26, 400.40) (561.26, 408.06) (559.03, 408.06)      /T1_2 .	<|special_separator|>
-(306.15, 389.65) (498.96, 389.65) (498.96, 397.31) (306.15, 397.31)      /T1_2 WeusedthepreprocesseddatafromLotfollahietal.	<|special_separator|>
-(498.88, 393.84) (501.79, 393.84) (501.79, 398.43) (498.88, 398.43)      /T1_2 8	<|special_separator|>
-(501.74, 389.65) (561.26, 389.65) (561.26, 397.31) (501.74, 397.31)      /T1_2 ,whichincludes	<|special_separator|>
-(306.15, 378.90) (331.41, 378.90) (331.41, 386.56) (306.15, 386.56)      /T1_2 library	<|special_separator|>
-(332.99, 378.90) (347.56, 378.90) (347.56, 386.56) (332.99, 386.56)      /T1_2 size	<|special_separator|>
-(349.14, 378.90) (405.07, 378.90) (405.07, 386.56) (349.14, 386.56)      /T1_2 normalization,	<|special_separator|>
-(406.65, 378.90) (419.93, 378.90) (419.93, 386.56) (406.65, 386.56)      /T1_2 cell	<|special_separator|>
-(421.51, 378.90) (435.78, 378.90) (435.78, 386.56) (421.51, 386.56)      /T1_2 and	<|special_separator|>
-(437.36, 378.90) (455.70, 378.90) (455.70, 386.56) (437.36, 386.56)      /T1_2 gene	<|special_separator|>
-(457.28, 378.90) (490.29, 378.90) (490.29, 386.56) (457.28, 386.56)      /T1_2 filtering,	<|special_separator|>
-(491.87, 378.90) (506.15, 378.90) (506.15, 386.56) (491.87, 386.56)      /T1_2 and	<|special_separator|>
-(507.73, 378.90) (512.00, 378.90) (512.00, 386.56) (507.73, 386.56)      /T1_2 a	<|special_separator|>
-(513.58, 379.22) (537.83, 379.22) (537.83, 385.33) (513.58, 385.33)      /T1_5 log1p	<|special_separator|>
-(539.29, 378.90) (561.26, 378.90) (561.26, 386.56) (539.29, 386.56)      /T1_2 trans-	<|special_separator|>
-(306.14, 368.15) (561.16, 368.15) (561.16, 375.81) (306.14, 375.81)      /T1_2 formation.Thedatasetconsistsof762,039single-cellmeasurements,	<|special_separator|>
-(306.14, 357.40) (561.23, 357.40) (561.23, 365.06) (306.14, 365.06)      /T1_2 outofwhich17,565belongtothecontrolpopulationand,onaverage,	<|special_separator|>
-(306.14, 346.65) (327.98, 346.65) (327.98, 354.31) (306.14, 354.31)      /T1_2 4,032	<|special_separator|>
-(329.80, 346.65) (381.04, 346.65) (381.04, 354.31) (329.80, 354.31)      /T1_2 observations	<|special_separator|>
-(382.86, 346.65) (391.02, 346.65) (391.02, 354.31) (382.86, 354.31)      /T1_2 of	<|special_separator|>
-(392.84, 346.65) (411.26, 346.65) (411.26, 354.31) (392.84, 354.31)      /T1_2 each	<|special_separator|>
-(413.07, 346.65) (431.92, 346.65) (431.92, 354.31) (413.07, 354.31)      /T1_2 drug	<|special_separator|>
-(433.74, 346.65) (448.45, 346.65) (448.45, 354.31) (433.74, 354.31)      /T1_2 and	<|special_separator|>
-(450.27, 346.65) (469.12, 346.65) (469.12, 354.31) (450.27, 354.31)      /T1_2 drug	<|special_separator|>
-(470.94, 346.65) (489.61, 346.65) (489.61, 354.31) (470.94, 354.31)      /T1_2 dose	<|special_separator|>
-(491.42, 346.65) (532.12, 346.65) (532.12, 354.31) (491.42, 354.31)      /T1_2 condition.	<|special_separator|>
-(533.94, 346.65) (561.27, 346.65) (561.27, 354.31) (533.94, 354.31)      /T1_2 During	<|special_separator|>
-(306.14, 335.90) (561.26, 335.90) (561.26, 343.56) (306.14, 343.56)      /T1_2 trainingandevaluation,weconsideredonlythe1,000highlyvariable	<|special_separator|>
-(306.14, 325.15) (436.15, 325.15) (436.15, 332.81) (306.14, 332.81)      /T1_2 genescomputedbyLotfollahietal.	<|special_separator|>
-(436.06, 329.34) (438.97, 329.34) (438.97, 333.93) (436.06, 333.93)      /T1_2 8	<|special_separator|>
-(438.97, 325.15) (561.27, 325.15) (561.27, 332.81) (438.97, 332.81)      /T1_2 .Asmanygeneswereunaffected,	<|special_separator|>
-(306.14, 314.40) (334.28, 314.40) (334.28, 322.06) (306.14, 322.06)      /T1_2 instead	<|special_separator|>
-(335.93, 314.40) (343.94, 314.40) (343.94, 322.06) (335.93, 322.06)      /T1_2 of	<|special_separator|>
-(345.60, 314.40) (385.79, 314.40) (385.79, 322.06) (345.60, 322.06)      /T1_2 evaluating	<|special_separator|>
-(387.44, 314.40) (394.81, 314.40) (394.81, 322.06) (387.44, 322.06)      /T1_2 in	<|special_separator|>
-(396.46, 314.40) (409.08, 314.40) (409.08, 322.06) (396.46, 322.06)      /T1_2 the	<|special_separator|>
-(410.73, 314.40) (483.03, 314.40) (483.03, 322.06) (410.73, 322.06)      /T1_2 1,000-dimensional	<|special_separator|>
-(484.68, 314.40) (503.22, 314.40) (503.22, 322.06) (484.68, 322.06)      /T1_2 gene	<|special_separator|>
-(504.88, 314.40) (561.21, 314.40) (561.21, 322.06) (504.88, 322.06)      /T1_2 space,weused	<|special_separator|>
-(306.14, 303.65) (561.17, 303.65) (561.17, 311.31) (306.14, 311.31)      /T1_2 onlythetop50differentiallyexpressedmarkergenesobtainedthrough	<|special_separator|>
-(306.14, 292.90) (324.53, 292.90) (324.53, 300.56) (306.14, 300.56)      /T1_2 gene	<|special_separator|>
-(326.14, 292.90) (355.36, 292.90) (355.36, 300.56) (326.14, 300.56)      /T1_2 ranking	<|special_separator|>
-(355.34, 297.09) (360.89, 297.09) (360.89, 301.68) (355.34, 301.68)      /T1_2 62	<|special_separator|>
-(360.90, 292.90) (363.13, 292.90) (363.13, 300.56) (360.90, 300.56)      /T1_2 .	<|special_separator|>
-(364.74, 292.90) (374.20, 292.90) (374.20, 300.56) (364.74, 300.56)      /T1_2 To	<|special_separator|>
-(375.82, 292.90) (409.38, 292.90) (409.38, 300.56) (375.82, 300.56)      /T1_2 facilitate	<|special_separator|>
-(410.99, 292.90) (443.52, 292.90) (443.52, 300.56) (410.99, 300.56)      /T1_2 training,	<|special_separator|>
-(445.13, 292.90) (488.72, 292.90) (488.72, 300.56) (445.13, 300.56)      /T1_2 wereduced	<|special_separator|>
-(490.33, 292.90) (502.85, 292.90) (502.85, 300.56) (490.33, 300.56)      /T1_2 the	<|special_separator|>
-(504.46, 292.90) (561.36, 292.90) (561.36, 300.56) (504.46, 300.56)      /T1_2 dimensionality	<|special_separator|>
-(306.14, 282.15) (540.40, 282.15) (540.40, 289.81) (306.14, 289.81)      /T1_2 ofthe1,000-dimensionalgeneexpressionfollowingBunneetal.	<|special_separator|>
-(540.54, 286.34) (546.48, 286.34) (546.48, 290.93) (540.54, 290.93)      /T1_2 30	<|special_separator|>
-(546.38, 282.15) (561.26, 282.15) (561.26, 289.81) (546.38, 289.81)      /T1_2 .We	<|special_separator|>
-(306.14, 271.40) (442.23, 271.40) (442.23, 279.06) (306.14, 279.06)      /T1_2 encodedgeneexpressiondataintoa	<|special_separator|>
-(443.64, 271.41) (447.98, 271.41) (447.98, 279.06) (443.64, 279.06)      /T1_3 k	<|special_separator|>
-(449.39, 271.40) (561.26, 271.40) (561.26, 279.06) (449.39, 279.06)      /T1_2 =50-dimensionallatentspace	<|special_separator|>
-(306.14, 260.65) (315.41, 260.65) (315.41, 268.31) (306.14, 268.31)      /T1_2 by	<|special_separator|>
-(317.01, 260.65) (347.18, 260.65) (347.18, 268.31) (317.01, 268.31)      /T1_2 training	<|special_separator|>
-(348.78, 260.65) (353.05, 260.65) (353.05, 268.31) (348.78, 268.31)      /T1_2 a	<|special_separator|>
-(354.65, 260.65) (379.50, 260.65) (379.50, 268.31) (354.65, 268.31)      /T1_2 vanilla	<|special_separator|>
-(381.10, 260.65) (429.83, 260.65) (429.83, 268.31) (381.10, 268.31)      /T1_2 autoencoder	<|special_separator|>
-(431.43, 260.65) (448.10, 260.65) (448.10, 268.31) (431.43, 268.31)      /T1_2 with	<|special_separator|>
-(449.70, 260.65) (458.90, 260.65) (458.90, 268.31) (449.70, 268.31)      /T1_2 an	<|special_separator|>
-(460.50, 260.65) (491.89, 260.65) (491.89, 268.31) (460.50, 268.31)      /T1_2 encoder	<|special_separator|>
-(494.26, 259.42) (498.27, 259.42) (498.27, 268.18) (494.26, 268.18)      /C0_7 E	<|special_separator|>
-(498.27, 258.72) (502.00, 258.72) (502.00, 264.86) (498.27, 264.86)      /C0_7 ϕ	<|special_separator|>
-(504.58, 259.41) (508.48, 259.41) (508.48, 268.20) (504.58, 268.20)      /C0_8 ∶	<|special_separator|>
-(510.55, 259.41) (516.00, 259.41) (516.00, 268.20) (510.55, 268.20)      /C0_8 ℝ	<|special_separator|>
-(516.01, 262.98) (519.07, 262.98) (519.07, 269.12) (516.01, 269.12)      /C0_7 d	<|special_separator|>
-(521.65, 259.41) (536.25, 259.41) (536.25, 268.20) (521.65, 268.20)      /C0_8 →ℝ	<|special_separator|>
-(536.25, 262.98) (539.00, 262.98) (539.00, 269.12) (536.25, 269.12)      /C0_7 k	<|special_separator|>
-(541.10, 260.65) (555.38, 260.65) (555.38, 268.31) (541.10, 268.31)      /T1_2 and	<|special_separator|>
-(556.98, 260.65) (561.26, 260.65) (561.26, 268.31) (556.98, 268.31)      /T1_2 a	<|special_separator|>
-(306.14, 249.90) (336.74, 249.90) (336.74, 257.56) (306.14, 257.56)      /T1_2 decoder	<|special_separator|>
-(338.39, 248.66) (343.64, 248.66) (343.64, 257.43) (338.39, 257.43)      /C0_9 D	<|special_separator|>
-(343.64, 247.97) (346.59, 247.97) (346.59, 254.11) (343.64, 254.11)      /C0_9 θ	<|special_separator|>
-(349.16, 248.65) (353.07, 248.65) (353.07, 257.45) (349.16, 257.45)     /C0_10 ∶	<|special_separator|>
-(355.14, 248.65) (360.58, 248.65) (360.58, 257.45) (355.14, 257.45)     /C0_10 ℝ	<|special_separator|>
-(360.59, 252.23) (363.34, 252.23) (363.34, 258.36) (360.59, 258.36)      /C0_9 k	<|special_separator|>
-(365.92, 248.65) (380.52, 248.65) (380.52, 257.45) (365.92, 257.45)     /C0_10 →ℝ	<|special_separator|>
-(380.52, 252.23) (383.59, 252.23) (383.59, 258.36) (380.52, 258.36)      /C0_9 d	<|special_separator|>
-(383.06, 249.90) (403.97, 249.90) (403.97, 257.56) (383.06, 257.56)      /T1_2 .Both	<|special_separator|>
-(404.92, 249.91) (409.26, 249.91) (409.26, 257.56) (404.92, 257.56)      /T1_3 E	<|special_separator|>
-(409.21, 248.93) (412.67, 248.93) (412.67, 253.52) (409.21, 253.52)      /T1_3 φ	<|special_separator|>
-(413.63, 249.90) (427.52, 249.90) (427.52, 257.56) (413.63, 257.56)      /T1_2 and	<|special_separator|>
-(428.47, 249.91) (434.15, 249.91) (434.15, 257.56) (428.47, 257.56)      /T1_3 D	<|special_separator|>
-(434.09, 248.93) (436.83, 248.93) (436.83, 253.52) (434.09, 253.52)      /T1_3 θ	<|special_separator|>
-(437.80, 249.90) (561.28, 249.90) (561.28, 257.56) (437.80, 257.56)      /T1_2 wereparameterizedbymulti-layer	<|special_separator|>
-(306.14, 239.15) (359.44, 239.15) (359.44, 246.81) (306.14, 246.81)      /T1_2 perceptrons.	<|special_separator|>
-(362.96, 239.15) (378.37, 239.15) (378.37, 246.81) (362.96, 246.81)      /T1_2 The	<|special_separator|>
-(381.89, 239.15) (406.46, 239.15) (406.46, 246.81) (381.89, 246.81)      /T1_2 entire	<|special_separator|>
-(409.98, 239.15) (462.86, 239.15) (462.86, 246.81) (409.98, 246.81)      /T1_2 autoencoder	<|special_separator|>
-(466.37, 239.15) (481.47, 239.15) (481.47, 246.81) (466.37, 246.81)      /T1_2 was	<|special_separator|>
-(484.99, 239.15) (527.31, 239.15) (527.31, 246.81) (484.99, 246.81)      /T1_2 optimized	<|special_separator|>
-(530.83, 239.15) (553.29, 239.15) (553.29, 246.81) (530.83, 246.81)      /T1_2 using	<|special_separator|>
-(556.81, 239.15) (561.17, 239.15) (561.17, 246.81) (556.81, 246.81)      /T1_2 a	<|special_separator|>
-(306.14, 228.40) (360.83, 228.40) (360.83, 236.06) (306.14, 236.06)      /T1_2 mean-squared	<|special_separator|>
-(362.46, 228.40) (382.14, 228.40) (382.14, 236.06) (362.46, 236.06)      /T1_2 error	<|special_separator|>
-(383.77, 228.40) (440.95, 228.40) (440.95, 236.06) (383.77, 236.06)      /T1_2 reconstruction	<|special_separator|>
-(442.58, 228.40) (459.64, 228.40) (459.64, 236.06) (442.58, 236.06)      /T1_2 loss.	<|special_separator|>
-(461.27, 228.40) (477.58, 228.40) (477.58, 236.06) (461.27, 236.06)      /T1_2 This	<|special_separator|>
-(479.21, 228.40) (528.29, 228.40) (528.29, 236.06) (479.21, 236.06)      /T1_2 autoencoder	<|special_separator|>
-(529.92, 228.40) (544.19, 228.40) (544.19, 236.06) (529.92, 236.06)      /T1_2 was	<|special_separator|>
-(545.82, 228.40) (561.26, 228.40) (561.26, 236.06) (545.82, 236.06)      /T1_2 pre-	<|special_separator|>
-(306.14, 217.65) (334.01, 217.65) (334.01, 225.31) (306.14, 225.31)      /T1_2 trained	<|special_separator|>
-(335.66, 217.65) (354.35, 217.65) (354.35, 225.31) (335.66, 225.31)      /T1_2 once	<|special_separator|>
-(356.00, 217.65) (366.01, 217.65) (366.01, 225.31) (356.00, 225.31)      /T1_2 on	<|special_separator|>
-(367.67, 217.65) (382.12, 217.65) (382.12, 225.31) (367.67, 225.31)      /T1_2 one	<|special_separator|>
-(383.78, 217.65) (405.46, 217.65) (405.46, 225.31) (383.78, 225.31)      /T1_2 batch	<|special_separator|>
-(407.11, 217.65) (415.12, 217.65) (415.12, 225.31) (407.11, 225.31)      /T1_2 of	<|special_separator|>
-(416.77, 217.65) (434.79, 217.65) (434.79, 225.31) (416.77, 225.31)      /T1_2 each	<|special_separator|>
-(436.45, 217.65) (476.18, 217.65) (476.18, 225.31) (436.45, 225.31)      /T1_2 condition,	<|special_separator|>
-(477.83, 217.65) (492.25, 217.65) (492.25, 225.31) (477.83, 225.31)      /T1_2 and	<|special_separator|>
-(493.91, 217.65) (499.49, 217.65) (499.49, 225.31) (493.91, 225.31)      /T1_2 it	<|special_separator|>
-(501.14, 217.65) (515.47, 217.65) (515.47, 225.31) (501.14, 225.31)      /T1_2 was	<|special_separator|>
-(517.12, 217.65) (534.72, 217.65) (534.72, 225.31) (517.12, 225.31)      /T1_2 then	<|special_separator|>
-(536.38, 217.65) (561.22, 217.65) (561.22, 225.31) (536.38, 225.31)      /T1_2 frozen	<|special_separator|>
-(306.14, 206.90) (331.67, 206.90) (331.67, 214.56) (306.14, 214.56)      /T1_2 during	<|special_separator|>
-(333.24, 206.90) (396.95, 206.90) (396.95, 214.56) (333.24, 214.56)      /T1_2 CMongetraining	<|special_separator|>
-(398.52, 206.90) (412.80, 206.90) (412.80, 214.56) (398.52, 214.56)      /T1_2 and	<|special_separator|>
-(414.37, 206.90) (456.48, 206.90) (456.48, 214.56) (414.37, 214.56)      /T1_2 evaluation.	<|special_separator|>
-(458.05, 206.90) (502.26, 206.90) (502.26, 214.56) (458.05, 214.56)      /T1_2 OTlearning	<|special_separator|>
-(503.84, 206.90) (538.36, 206.90) (538.36, 214.56) (503.84, 214.56)      /T1_2 occurred	<|special_separator|>
-(539.93, 206.90) (547.22, 206.90) (547.22, 214.56) (539.93, 214.56)      /T1_2 in	<|special_separator|>
-(548.79, 206.90) (561.27, 206.90) (561.27, 214.56) (548.79, 214.56)      /T1_2 the	<|special_separator|>
-(306.14, 196.15) (328.32, 196.15) (328.32, 203.81) (306.14, 203.81)      /T1_2 latent	<|special_separator|>
-(329.82, 196.15) (351.42, 196.15) (351.42, 203.81) (329.82, 203.81)      /T1_2 space	<|special_separator|>
-(352.92, 196.15) (360.85, 196.15) (360.85, 203.81) (352.92, 203.81)      /T1_2 of	<|special_separator|>
-(362.35, 196.15) (374.83, 196.15) (374.83, 203.81) (362.35, 203.81)      /T1_2 the	<|special_separator|>
-(376.33, 196.15) (426.97, 196.15) (426.97, 203.81) (376.33, 203.81)      /T1_2 autoencoder.	<|special_separator|>
-(428.47, 196.15) (478.28, 196.15) (478.28, 203.81) (428.47, 203.81)      /T1_2 Thereported	<|special_separator|>
-(479.78, 196.15) (508.58, 196.15) (508.58, 203.81) (479.78, 203.81)      /T1_2 metrics	<|special_separator|>
-(510.08, 196.15) (513.12, 196.15) (513.12, 203.81) (510.08, 203.81)      /T1_2 (	<|special_separator|>
-(513.12, 196.16) (518.27, 196.16) (518.27, 203.81) (513.12, 203.81)      /T1_3 R	<|special_separator|>
-(518.24, 200.34) (520.92, 200.34) (520.92, 204.93) (518.24, 204.93)      /T1_2 2	<|special_separator|>
-(520.92, 196.15) (561.27, 196.15) (561.27, 203.81) (520.92, 203.81)      /T1_2 ,MMDand	<|special_separator|>
-(306.14, 185.40) (355.62, 185.40) (355.62, 193.06) (306.14, 193.06)      /T1_2 Wasserstein)	<|special_separator|>
-(357.33, 185.40) (375.91, 185.40) (375.91, 193.06) (357.33, 193.06)      /T1_2 were	<|special_separator|>
-(377.62, 185.40) (417.76, 185.40) (417.76, 193.06) (377.62, 193.06)      /T1_2 calculated	<|special_separator|>
-(419.47, 185.40) (426.90, 185.40) (426.90, 193.06) (419.47, 193.06)      /T1_2 in	<|special_separator|>
-(428.61, 185.40) (441.36, 185.40) (441.36, 193.06) (428.61, 193.06)      /T1_2 the	<|special_separator|>
-(443.07, 185.40) (473.29, 185.40) (473.29, 193.06) (443.07, 193.06)      /T1_2 original	<|special_separator|>
-(475.00, 185.40) (493.72, 185.40) (493.72, 193.06) (475.00, 193.06)      /T1_2 gene	<|special_separator|>
-(495.44, 185.40) (537.64, 185.40) (537.64, 193.06) (495.44, 193.06)      /T1_2 expression	<|special_separator|>
-(539.35, 185.40) (561.42, 185.40) (561.42, 193.06) (539.35, 193.06)      /T1_2 space	<|special_separator|>
-(306.14, 174.65) (535.02, 174.65) (535.02, 182.31) (306.14, 182.31)      /T1_2 (betweenmeasuredgeneexpressionanddecodedtransport).	<|special_separator|>
-(323.15, 163.90) (354.42, 163.90) (354.42, 171.56) (323.15, 171.56)      /T1_2 Second,	<|special_separator|>
-(356.16, 163.90) (368.97, 163.90) (368.97, 171.56) (356.16, 171.56)      /T1_2 the	<|special_separator|>
-(370.70, 163.90) (378.11, 163.90) (378.11, 171.56) (370.70, 171.56)      /T1_2 4i	<|special_separator|>
-(379.84, 163.90) (408.65, 163.90) (408.65, 171.56) (379.84, 171.56)      /T1_2 dataset	<|special_separator|>
-(410.39, 163.90) (443.90, 163.90) (443.90, 171.56) (410.39, 171.56)      /T1_2 contains	<|special_separator|>
-(445.63, 163.90) (492.74, 163.90) (492.74, 171.56) (445.63, 171.56)      /T1_2 multiplexed	<|special_separator|>
-(494.48, 163.90) (523.45, 163.90) (523.45, 171.56) (494.48, 171.56)      /T1_2 protein	<|special_separator|>
-(525.18, 163.90) (561.18, 163.90) (561.18, 171.56) (525.18, 171.56)      /T1_2 measure-	<|special_separator|>
-(306.14, 153.15) (329.98, 153.15) (329.98, 160.81) (306.14, 160.81)      /T1_2 ments	<|special_separator|>
-(331.61, 153.15) (339.59, 153.15) (339.59, 160.81) (331.61, 160.81)      /T1_2 of	<|special_separator|>
-(341.22, 153.15) (386.77, 153.15) (386.77, 160.81) (341.22, 160.81)      /T1_2 40different	<|special_separator|>
-(388.41, 153.15) (420.52, 153.15) (420.52, 160.81) (388.41, 160.81)      /T1_2 proteins	<|special_separator|>
-(422.16, 153.15) (459.95, 153.15) (459.95, 160.81) (422.16, 160.81)      /T1_2 measured	<|special_separator|>
-(461.58, 153.15) (468.91, 153.15) (468.91, 160.81) (461.58, 160.81)      /T1_2 in	<|special_separator|>
-(470.55, 153.15) (494.92, 153.15) (494.92, 160.81) (470.55, 160.81)      /T1_2 97,748	<|special_separator|>
-(496.55, 153.15) (513.70, 153.15) (513.70, 160.81) (496.55, 160.81)      /T1_2 cells	<|special_separator|>
-(515.33, 153.15) (542.92, 153.15) (542.92, 160.81) (515.33, 160.81)      /T1_2 (10,995	<|special_separator|>
-(544.55, 153.15) (561.26, 153.15) (561.26, 160.81) (544.55, 160.81)      /T1_2 con-	<|special_separator|>
-(306.14, 142.40) (327.53, 142.40) (327.53, 150.06) (306.14, 150.06)      /T1_2 trols)	<|special_separator|>
-(329.35, 142.40) (348.70, 142.40) (348.70, 150.06) (329.35, 150.06)      /T1_2 from	<|special_separator|>
-(350.53, 142.40) (365.25, 142.40) (365.25, 150.06) (350.53, 150.06)      /T1_2 two	<|special_separator|>
-(367.07, 142.40) (385.50, 142.40) (385.50, 150.06) (367.07, 150.06)      /T1_2 lines	<|special_separator|>
-(387.33, 142.40) (395.49, 142.40) (395.49, 150.06) (387.33, 150.06)      /T1_2 of	<|special_separator|>
-(397.32, 142.40) (438.82, 142.40) (438.82, 150.06) (397.32, 150.06)      /T1_2 melanoma	<|special_separator|>
-(440.64, 142.40) (474.44, 142.40) (474.44, 150.06) (440.64, 150.06)      /T1_2 tumours	<|special_separator|>
-(476.26, 142.40) (504.96, 142.40) (504.96, 150.06) (476.26, 150.06)      /T1_2 treated	<|special_separator|>
-(506.78, 142.40) (524.02, 142.40) (524.02, 150.06) (506.78, 150.06)      /T1_2 with	<|special_separator|>
-(525.84, 142.40) (540.59, 142.40) (540.59, 150.06) (525.84, 150.06)      /T1_2 one	<|special_separator|>
-(542.41, 142.40) (550.58, 142.40) (550.58, 150.06) (542.41, 150.06)      /T1_2 of	<|special_separator|>
-(552.40, 142.40) (561.31, 142.40) (561.31, 150.06) (552.40, 150.06)      /T1_2 35	<|special_separator|>
-(306.14, 131.65) (470.98, 131.65) (470.98, 139.31) (306.14, 139.31)      /T1_2 cancertherapies,eachinvolving~2,500cells	<|special_separator|>
-(470.97, 135.84) (476.63, 135.84) (476.63, 140.43) (470.97, 140.43)      /T1_2 36	<|special_separator|>
-(476.55, 131.65) (561.27, 131.65) (561.27, 139.31) (476.55, 139.31)      /T1_2 .Weobtainedpreproc-	<|special_separator|>
-(306.14, 120.90) (327.92, 120.90) (327.92, 128.56) (306.14, 128.56)      /T1_2 essed	<|special_separator|>
-(329.66, 120.90) (346.92, 120.90) (346.92, 128.56) (329.66, 128.56)      /T1_2 data	<|special_separator|>
-(348.66, 120.90) (367.90, 120.90) (367.90, 128.56) (348.66, 128.56)      /T1_2 from	<|special_separator|>
-(369.65, 120.90) (382.66, 120.90) (382.66, 128.56) (369.65, 128.56)      /T1_2 ref.	<|special_separator|>
-(384.40, 120.90) (395.77, 120.90) (395.77, 128.56) (384.40, 128.56)      /T1_2 33,	<|special_separator|>
-(397.51, 120.90) (432.52, 120.90) (432.52, 128.56) (397.51, 128.56)      /T1_2 resulting	<|special_separator|>
-(434.26, 120.90) (441.74, 120.90) (441.74, 128.56) (434.26, 128.56)      /T1_2 in	<|special_separator|>
-(443.48, 120.90) (453.35, 120.90) (453.35, 128.56) (443.48, 128.56)      /T1_2 48	<|special_separator|>
-(455.09, 120.90) (487.07, 120.90) (487.07, 128.56) (455.09, 128.56)      /T1_2 features	<|special_separator|>
-(488.81, 120.90) (500.52, 120.90) (500.52, 128.56) (488.81, 128.56)      /T1_2 for	<|special_separator|>
-(502.26, 120.90) (531.45, 120.90) (531.45, 128.56) (502.26, 128.56)      /T1_2 cellular	<|special_separator|>
-(533.20, 120.90) (561.25, 120.90) (561.25, 128.56) (533.20, 128.56)      /T1_2 marker	<|special_separator|>
-(306.14, 110.15) (561.39, 110.15) (561.39, 117.81) (306.14, 117.81)      /T1_2 expressionandcellshape.Weexcludedtwooftheeightcombinatorial	<|special_separator|>
-(306.14, 099.40) (349.14, 099.40) (349.14, 107.06) (306.14, 107.06)      /T1_2 treatments	<|special_separator|>
-(350.86, 099.40) (382.47, 099.40) (382.47, 107.06) (350.86, 107.06)      /T1_2 because	<|special_separator|>
-(384.18, 099.40) (401.35, 099.40) (401.35, 107.06) (384.18, 107.06)      /T1_2 they	<|special_separator|>
-(403.07, 099.40) (415.85, 099.40) (415.85, 107.06) (403.07, 107.06)      /T1_2 did	<|special_separator|>
-(417.56, 099.40) (430.92, 099.40) (430.92, 107.06) (417.56, 107.06)      /T1_2 not	<|special_separator|>
-(432.64, 099.40) (454.91, 099.40) (454.91, 107.06) (432.64, 107.06)      /T1_2 occur	<|special_separator|>
-(456.62, 099.40) (464.80, 099.40) (464.80, 107.06) (456.62, 107.06)      /T1_2 as	<|special_separator|>
-(466.51, 099.40) (470.84, 099.40) (470.84, 107.06) (466.51, 107.06)      /T1_2 a	<|special_separator|>
-(472.56, 099.40) (495.69, 099.40) (495.69, 107.06) (472.56, 107.06)      /T1_2 single	<|special_separator|>
-(497.40, 099.40) (538.92, 099.40) (538.92, 107.06) (497.40, 107.06)      /T1_2 treatment,	<|special_separator|>
-(540.63, 099.40) (561.23, 099.40) (561.23, 107.06) (540.63, 107.06)      /T1_2 aswe	<|special_separator|>
-(306.14, 088.65) (519.60, 088.65) (519.60, 096.31) (306.14, 096.31)      /T1_2 basedtheMoAdrugembeddingonthesingletreatments.	<|special_separator|>
-(306.14, 067.19) (445.25, 067.19) (445.25, 075.17) (306.14, 075.17)      /T1_4 Hyperparametersandmodelsizes	<|special_separator|>
-(306.14, 056.40) (313.82, 056.40) (313.82, 064.06) (306.14, 064.06)      /T1_2 In	<|special_separator|>
-(315.71, 056.40) (325.04, 056.40) (325.04, 064.06) (315.71, 064.06)      /T1_2 all	<|special_separator|>
-(326.93, 056.40) (378.91, 056.40) (378.91, 064.06) (326.93, 064.06)      /T1_2 experiments,	<|special_separator|>
-(380.80, 056.40) (391.61, 056.40) (391.61, 064.06) (380.80, 064.06)      /T1_2 we	<|special_separator|>
-(393.49, 056.40) (412.18, 056.40) (412.18, 064.06) (393.49, 064.06)      /T1_2 used	<|special_separator|>
-(414.07, 056.40) (426.98, 056.40) (426.98, 064.06) (414.07, 064.06)      /T1_2 the	<|special_separator|>
-(428.87, 056.40) (500.39, 056.40) (500.39, 064.06) (428.87, 064.06)      /T1_2 AdamWoptimizer	<|special_separator|>
-(500.41, 060.59) (506.17, 060.59) (506.17, 065.18) (500.41, 065.18)      /T1_2 63	<|special_separator|>
-(508.01, 056.40) (525.27, 056.40) (525.27, 064.06) (508.01, 064.06)      /T1_2 with	<|special_separator|>
-(527.16, 056.40) (536.63, 056.40) (536.63, 064.06) (527.16, 064.06)      /T1_2 an	<|special_separator|>
-(538.51, 056.40) (561.29, 056.40) (561.29, 064.06) (538.51, 064.06)      /T1_2 initial	<|special_separator|>
-(306.14, 045.65) (373.35, 045.65) (373.35, 053.31) (306.14, 053.31)      /T1_2 learningrateof10	<|special_separator|>
-(373.34, 049.84) (379.06, 049.84) (379.06, 054.43) (373.34, 054.43)      /T1_2 -4	<|special_separator|>
-(380.37, 045.65) (510.43, 045.65) (510.43, 053.31) (380.37, 053.31)      /T1_2 andweightdecayregularization10	<|special_separator|>
-(510.42, 049.84) (515.71, 049.84) (515.71, 054.43) (510.42, 054.43)      /T1_2 -5	<|special_separator|>
-(515.58, 045.65) (561.26, 045.65) (561.26, 053.31) (515.58, 053.31)      /T1_2 .TheMonge

groundtruth/parser/font_11.pdf.page_no_11.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:00c479adf07606021e35f0f3949f54f7ea87a2da154f4e3c8bb5c1db22271685
-size 5370187

groundtruth/parser/font_11.pdf.page_no_11.py.json.line.txt

@@ -1,217 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence | Volume 8 | June 2026 | 984-996	<|special_separator|>
-(546.19, 020.07) (562.05, 020.07) (562.05, 027.05) (546.19, 027.05)      /T1_1 994	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 733.65) (296.23, 733.65) (296.23, 741.31) (039.69, 741.31)      /T1_2 network has 4 hidden layers with 64 neurons each. The dose and drug 	<|special_separator|>
-(039.69, 722.90) (296.19, 722.90) (296.19, 730.56) (039.69, 730.56)      /T1_2 embedders were parameterized with one dense layer. The Euclidean 	<|special_separator|>
-(039.69, 712.15) (294.82, 712.15) (294.82, 719.81) (039.69, 719.81)      /T1_2 distance was used as the displacement cost, and the Monge Gap regu-	<|special_separator|>
-(039.69, 701.40) (065.42, 701.40) (065.42, 709.06) (039.69, 709.06)      /T1_2 larizer 	<|special_separator|>
-(065.04, 701.41) (069.24, 701.41) (069.24, 709.06) (065.04, 709.06)      /T1_3 λ	<|special_separator|>
-(069.24, 701.40) (085.01, 701.40) (085.01, 709.06) (069.24, 709.06)      /T1_2  = 10	<|special_separator|>
-(085.00, 705.59) (090.38, 705.59) (090.38, 710.18) (085.00, 710.18)      /T1_2 -2	<|special_separator|>
-(090.38, 701.40) (296.35, 701.40) (296.35, 709.06) (090.38, 709.06)      /T1_2 . During the OT training phase, we repeatedly sampled a 	<|special_separator|>
-(039.69, 690.65) (296.16, 690.65) (296.16, 698.31) (039.69, 698.31)      /T1_2 batch of 256 observations from the source and target distributions for 	<|special_separator|>
-(039.69, 679.90) (296.27, 679.90) (296.27, 687.56) (039.69, 687.56)      /T1_2 1,000 iterations for local models (without a condition or conditioned 	<|special_separator|>
-(039.69, 669.15) (296.16, 669.15) (296.16, 676.81) (039.69, 676.81)      /T1_2 on dose), for 10,000 iterations for the global models (RDKit and MoA) 	<|special_separator|>
-(039.69, 658.40) (296.22, 658.40) (296.22, 666.06) (039.69, 666.06)      /T1_2 and for 500,000 iterations for the bigger models that were trained 	<|special_separator|>
-(039.69, 647.65) (296.33, 647.65) (296.33, 655.31) (039.69, 655.31)      /T1_2 on all drugs in the SciPlex dataset. Each batch contained samples 	<|special_separator|>
-(039.69, 636.90) (296.24, 636.90) (296.24, 644.56) (039.69, 644.56)      /T1_2 from only one context, which was uniformly sampled. For the SciPlex 	<|special_separator|>
-(039.69, 626.15) (296.20, 626.15) (296.20, 633.81) (039.69, 633.81)      /T1_2 experiments, the autoencoder consisted of two hidden layers, with 	<|special_separator|>
-(039.69, 615.40) (296.23, 615.40) (296.23, 623.06) (039.69, 623.06)      /T1_2 512 dimensions in both the encoder and decoder. The 50-dimensional 	<|special_separator|>
-(039.69, 604.65) (296.24, 604.65) (296.24, 612.31) (039.69, 612.31)      /T1_2 latent representation was learned through 50 epochs with a batch size 	<|special_separator|>
-(039.69, 593.90) (214.10, 593.90) (214.10, 601.56) (039.69, 601.56)      /T1_2 of 256. All models were implemented using the 	<|special_separator|>
-(213.89, 594.22) (247.80, 594.22) (247.80, 600.33) (213.89, 600.33)      /T1_4 OTT-JAX	<|special_separator|>
-(247.67, 593.90) (280.14, 593.90) (280.14, 601.56) (247.67, 601.56)      /T1_2  package	<|special_separator|>
-(280.11, 598.09) (286.03, 598.09) (286.03, 602.68) (280.11, 602.68)      /T1_2 64	<|special_separator|>
-(285.96, 593.90) (288.18, 593.90) (288.18, 601.56) (285.96, 601.56)      /T1_2 .	<|special_separator|>
-(056.69, 583.15) (296.24, 583.15) (296.24, 590.81) (056.69, 590.81)      /T1_2 For the large-scale SciPlex experiment (712 training conditions), 	<|special_separator|>
-(039.69, 572.40) (296.26, 572.40) (296.26, 580.06) (039.69, 580.06)      /T1_2 we increased the model size of CMonge and the embedding size of the 	<|special_separator|>
-(039.69, 561.65) (296.25, 561.65) (296.25, 569.31) (039.69, 569.31)      /T1_2 gene expression and the context. The drug and data embedding was 	<|special_separator|>
-(039.69, 550.90) (296.29, 550.90) (296.29, 558.56) (039.69, 558.56)      /T1_2 increased from 50 to 100 dimensions, which increased the number of 	<|special_separator|>
-(039.69, 540.15) (296.26, 540.15) (296.26, 547.81) (039.69, 547.81)      /T1_2 parameters in the gene expression autoencoder from 1.60 million to 	<|special_separator|>
-(039.69, 529.40) (296.20, 529.40) (296.20, 537.06) (039.69, 537.06)      /T1_2 1.65 million. The four fully connected layers of CMonge were increased 	<|special_separator|>
-(039.69, 518.65) (296.32, 518.65) (296.32, 526.31) (039.69, 526.31)      /T1_2 from [64, 64, 64, 64] to [256, 256, 512, 512], resulting in a parameter 	<|special_separator|>
-(039.69, 507.90) (296.19, 507.90) (296.19, 515.56) (039.69, 515.56)      /T1_2 increase from 23,000 to 560,000 (580,000) for MoA embedding 	<|special_separator|>
-(039.69, 497.15) (296.40, 497.15) (296.40, 504.81) (039.69, 504.81)      /T1_2 (RDKit embedding, respectively). For comparison, chemCPA, which 	<|special_separator|>
-(039.69, 486.40) (294.80, 486.40) (294.80, 494.06) (039.69, 494.06)      /T1_2 has one model for embedding the data, context information and learn-	<|special_separator|>
-(039.69, 475.65) (251.63, 475.65) (251.63, 483.31) (039.69, 483.31)      /T1_2 ing the perturbation, has around 1.37 million parameters.	<|special_separator|>
-(039.69, 454.19) (100.40, 454.19) (100.40, 462.17) (039.69, 462.17)      /T1_5 Benchmarking	<|special_separator|>
-(039.69, 443.40) (294.80, 443.40) (294.80, 451.06) (039.69, 451.06)      /T1_2 We first benchmarked different state-of-the-art methods for uncondi-	<|special_separator|>
-(039.69, 432.65) (154.31, 432.65) (154.31, 440.31) (039.69, 440.31)      /T1_2 tional perturbation modelling.	<|special_separator|>
-(039.69, 411.53) (080.88, 411.53) (080.88, 418.75) (039.69, 418.75)      /T1_6 4i dataset	<|special_separator|>
-(080.97, 411.29) (083.46, 411.29) (083.46, 418.81) (080.97, 418.81)      /T1_5 .	<|special_separator|>
-(083.56, 411.15) (296.30, 411.15) (296.30, 418.81) (083.56, 418.81)      /T1_2  For the non-conditional experiments, we trained each 	<|special_separator|>
-(039.69, 400.40) (296.23, 400.40) (296.23, 408.06) (039.69, 408.06)      /T1_2 method on each of the 35 therapies with an 80/20 training/validation 	<|special_separator|>
-(039.69, 389.65) (176.62, 389.65) (176.62, 397.31) (039.69, 397.31)      /T1_2 split. We note that the original scGen	<|special_separator|>
-(176.60, 393.84) (179.52, 393.84) (179.52, 398.43) (176.60, 398.43)      /T1_2 8	<|special_separator|>
-(179.52, 389.65) (294.82, 389.65) (294.82, 397.31) (179.52, 397.31)      /T1_2  relies on a variational formula-	<|special_separator|>
-(039.69, 378.90) (055.31, 378.90) (055.31, 386.56) (039.69, 386.56)      /T1_2 tion	<|special_separator|>
-(055.35, 383.09) (060.85, 383.09) (060.85, 387.68) (055.35, 387.68)      /T1_2 65	<|special_separator|>
-(060.74, 378.90) (296.17, 378.90) (296.17, 386.56) (060.74, 386.56)      /T1_2 . Instead, in our experiments, we followed the set-up in Bunne 	<|special_separator|>
-(039.69, 368.15) (057.79, 368.15) (057.79, 375.81) (039.69, 375.81)      /T1_2 et al.	<|special_separator|>
-(057.81, 372.34) (063.17, 372.34) (063.17, 376.93) (057.81, 376.93)      /T1_2 33	<|special_separator|>
-(063.11, 368.15) (296.13, 368.15) (296.13, 375.81) (063.11, 375.81)      /T1_2 . That is, we used a vanilla autoencoder, and both the encoder 	<|special_separator|>
-(039.69, 357.40) (296.27, 357.40) (296.27, 365.06) (039.69, 365.06)      /T1_2 and the decoder were parameterized with fully connected layers. The 	<|special_separator|>
-(039.69, 346.65) (296.14, 346.65) (296.14, 354.31) (039.69, 354.31)      /T1_2 results in Supplementary Table 2 and Extended Data Fig. 1 confirm 	<|special_separator|>
-(039.69, 335.90) (296.25, 335.90) (296.25, 343.56) (039.69, 343.56)      /T1_2 the finding by ref. 32, namely that Monge Gap achieves overall the 	<|special_separator|>
-(039.69, 325.15) (296.27, 325.15) (296.27, 332.81) (039.69, 332.81)      /T1_2 best result with respect to each of the evaluation metrics and also has 	<|special_separator|>
-(039.69, 314.40) (296.36, 314.40) (296.36, 322.06) (039.69, 322.06)      /T1_2 a lower standard deviation. In Extended Data Fig. 1, the subplot for 	<|special_separator|>
-(039.69, 303.65) (296.25, 303.65) (296.25, 311.31) (039.69, 311.31)      /T1_2 the Wasserstein distance clearly shows that the Monge model (which 	<|special_separator|>
-(039.69, 292.90) (296.23, 292.90) (296.23, 300.56) (039.69, 300.56)      /T1_2 directly optimizes this metric) performed consistently, regardless of 	<|special_separator|>
-(039.69, 282.15) (296.28, 282.15) (296.28, 289.81) (039.69, 289.81)      /T1_2 the perturbation, whereas the autoencoder struggled to capture the 	<|special_separator|>
-(039.69, 271.40) (296.35, 271.40) (296.35, 279.06) (039.69, 279.06)      /T1_2 effects of the perturbation in some cases. On the other hand, the ICNN 	<|special_separator|>
-(039.69, 260.65) (296.38, 260.65) (296.38, 268.31) (039.69, 268.31)      /T1_2 results are skewed by a single outlier. Notably, the OT-based models 	<|special_separator|>
-(039.69, 249.90) (191.99, 249.90) (191.99, 257.56) (039.69, 257.56)      /T1_2 outperformed the autoencoder on MMD, 	<|special_separator|>
-(191.61, 249.91) (196.72, 249.91) (196.72, 257.56) (191.61, 257.56)      /T1_3 R	<|special_separator|>
-(196.70, 254.09) (199.36, 254.09) (199.36, 258.68) (196.70, 258.68)      /T1_2 2	<|special_separator|>
-(199.35, 249.90) (296.25, 249.90) (296.25, 257.56) (199.35, 257.56)      /T1_2  and drug signatures, even 	<|special_separator|>
-(039.69, 239.15) (282.05, 239.15) (282.05, 246.81) (039.69, 246.81)      /T1_2 though they were trained to optimize the primal and dual OT loss.	<|special_separator|>
-(039.69, 218.03) (068.61, 218.03) (068.61, 225.25) (039.69, 225.25)      /T1_6 SciPlex	<|special_separator|>
-(068.56, 217.79) (070.97, 217.79) (070.97, 225.31) (068.56, 225.31)      /T1_5 .	<|special_separator|>
-(070.95, 217.65) (296.22, 217.65) (296.22, 225.31) (070.95, 225.31)      /T1_2  For each of the nine drugs and each of the four doses, we fitted 	<|special_separator|>
-(039.69, 206.90) (296.22, 206.90) (296.22, 214.56) (039.69, 214.56)      /T1_2 a different model, resulting in 36 models per method. We included the 	<|special_separator|>
-(039.69, 196.15) (296.36, 196.15) (296.36, 203.81) (039.69, 203.81)      /T1_2 identity mapping as a baseline, which simply predicts the unperturbed 	<|special_separator|>
-(039.69, 185.40) (296.32, 185.40) (296.32, 193.06) (039.69, 193.06)      /T1_2 cell states. Supplementary Table 1 and Supplementary Fig. 4 show the 	<|special_separator|>
-(039.69, 174.65) (296.22, 174.65) (296.22, 182.31) (039.69, 182.31)      /T1_2 performance of the different methods. Our main evaluation metric 	<|special_separator|>
-(039.69, 163.90) (296.25, 163.90) (296.25, 171.56) (039.69, 171.56)      /T1_2 was still the coefficient of determination of the feature means for the 	<|special_separator|>
-(039.69, 153.15) (122.53, 153.15) (122.53, 160.81) (039.69, 160.81)      /T1_2 highly variable genes (	<|special_separator|>
-(122.53, 153.16) (127.68, 153.16) (127.68, 160.81) (122.53, 160.81)      /T1_3 R	<|special_separator|>
-(127.68, 157.34) (130.36, 157.34) (130.36, 161.93) (127.68, 161.93)      /T1_2 2	<|special_separator|>
-(130.36, 153.15) (296.26, 153.15) (296.26, 160.81) (130.36, 160.81)      /T1_2 ), but we also report the entropy-regularized 	<|special_separator|>
-(039.69, 142.40) (296.19, 142.40) (296.19, 150.06) (039.69, 150.06)      /T1_2 Wasserstein distance, which is closely related to the objective function 	<|special_separator|>
-(039.69, 131.65) (296.30, 131.65) (296.30, 139.31) (039.69, 139.31)      /T1_2 of the ICNN and Monge models. The ICNN and Monge models used the 	<|special_separator|>
-(039.69, 120.90) (285.89, 120.90) (285.89, 128.56) (039.69, 128.56)      /T1_2 50-dimensional latent representation learned by the autoencoder.	<|special_separator|>
-(056.69, 110.15) (296.15, 110.15) (296.15, 117.81) (056.69, 117.81)      /T1_2 All model predictions were decoded and evaluated in the cell space 	<|special_separator|>
-(039.69, 099.40) (296.36, 099.40) (296.36, 107.06) (039.69, 107.06)      /T1_2 for the 50 highly variable genes. We observed that the neural-OT-based 	<|special_separator|>
-(039.69, 088.65) (273.44, 088.65) (273.44, 096.31) (039.69, 096.31)      /T1_2 solvers significantly outperformed the autoencoder approach.	<|special_separator|>
-(056.69, 077.90) (296.31, 077.90) (296.31, 085.56) (056.69, 085.56)      /T1_2 Supplementary Fig. 5 shows that, although the ICNN-based solver 	<|special_separator|>
-(039.69, 067.15) (296.36, 067.15) (296.36, 074.81) (039.69, 074.81)      /T1_2 slightly outperformed the Monge-based counterpart, their results 	<|special_separator|>
-(039.69, 056.40) (294.78, 056.40) (294.78, 064.06) (039.69, 064.06)      /T1_2 were highly correlated. Moreover, they attained almost identical per-	<|special_separator|>
-(039.69, 045.65) (296.23, 045.65) (296.23, 053.31) (039.69, 053.31)      /T1_2 formance for the Wasserstein distance. Because of this and because 	<|special_separator|>
-(306.14, 733.65) (562.80, 733.65) (562.80, 741.31) (306.14, 741.31)      /T1_2 we are expanding upon the Monge-based methodology, we did not 	<|special_separator|>
-(306.14, 722.90) (493.68, 722.90) (493.68, 730.56) (306.14, 730.56)      /T1_2 include ICNN-based benchmarks in the main work.	<|special_separator|>
-(323.15, 712.15) (562.85, 712.15) (562.85, 719.81) (323.15, 719.81)      /T1_2 We additionally tested vanilla attention pooling and multi-head 	<|special_separator|>
-(306.14, 701.40) (561.26, 701.40) (561.26, 709.06) (306.14, 709.06)      /T1_2 attention pooling with the 4i data as the number of combination con-	<|special_separator|>
-(306.14, 690.65) (562.66, 690.65) (562.66, 698.31) (306.14, 698.31)      /T1_2 ditions as small. As we show in Supplementary Table 9, this yielded 	<|special_separator|>
-(306.14, 679.90) (561.22, 679.90) (561.22, 687.56) (306.14, 687.56)      /T1_2 marginal benefits over the 4i dataset only upon careful regularization.	<|special_separator|>
-(306.14, 658.44) (388.66, 658.44) (388.66, 666.42) (306.14, 666.42)      /T1_5 Reporting summary	<|special_separator|>
-(306.14, 647.65) (562.87, 647.65) (562.87, 655.31) (306.14, 655.31)      /T1_2 Further information on research design is available in the Nature 	<|special_separator|>
-(306.14, 636.90) (494.72, 636.90) (494.72, 644.56) (306.14, 644.56)      /T1_2 Portfolio Reporting Summary linked to this article.	<|special_separator|>
-(306.14, 615.10) (389.87, 615.10) (389.87, 624.86) (306.14, 624.86)      /T1_7 Data availability	<|special_separator|>
-(306.14, 604.65) (561.26, 604.65) (561.26, 612.31) (306.14, 612.31)      /T1_2 Both datasets used in this study are publicly available. The preproc-	<|special_separator|>
-(306.14, 593.90) (562.72, 593.90) (562.72, 601.56) (306.14, 601.56)      /T1_2 essed versions of the SciPlex3 and 4i datasets have been made available 	<|special_separator|>
-(306.14, 583.15) (561.25, 583.15) (561.25, 590.81) (306.14, 590.81)      /T1_2 under a permanent DOI and can be downloaded from http://www.	<|special_separator|>
-(306.14, 572.40) (561.20, 572.40) (561.20, 580.06) (306.14, 580.06)      /T1_2 research-collection.ethz.ch/entities/researchdata/7964ba4a-92c7-	<|special_separator|>
-(306.14, 561.65) (433.18, 561.65) (433.18, 569.31) (306.14, 569.31)      /T1_2 4031-80f4-358f681c5056 (ref. 66).	<|special_separator|>
-(306.14, 539.85) (391.75, 539.85) (391.75, 549.61) (306.14, 549.61)      /T1_7 Code availability	<|special_separator|>
-(306.14, 529.40) (562.72, 529.40) (562.72, 537.06) (306.14, 537.06)      /T1_2 The source code for reproducing the experiments is available via 	<|special_separator|>
-(306.14, 518.65) (562.74, 518.65) (562.74, 526.31) (306.14, 526.31)      /T1_2 Zenodo at https://doi.org/10.5281/zenodo.17854338 (ref. 67). It is 	<|special_separator|>
-(306.14, 507.90) (561.28, 507.90) (561.28, 515.56) (306.14, 515.56)      /T1_2 also available under an MIT licence via GitHub at https://github.com/	<|special_separator|>
-(306.14, 497.15) (561.31, 497.15) (561.31, 504.81) (306.14, 504.81)      /T1_2 AI4SCR/Conditional-Monge and can be installed from PyPI (http://	<|special_separator|>
-(306.14, 486.40) (451.90, 486.40) (451.90, 494.06) (306.14, 494.06)      /T1_2 pypi.org/project/cmonge/) as cmonge.	<|special_separator|>
-(306.14, 464.60) (362.82, 464.60) (362.82, 474.36) (306.14, 474.36)      /T1_7 References	<|special_separator|>
-(306.14, 454.56) (313.35, 454.56) (313.35, 461.57) (306.14, 461.57)      /T1_8 1. 	<|special_separator|>
-(323.15, 454.56) (556.70, 454.56) (556.70, 461.57) (323.15, 461.57)      /T1_8 Frangieh, C. J. et al. Multimodal pooled perturb-cite-seq screens 	<|special_separator|>
-(323.15, 443.80) (560.27, 443.80) (560.27, 450.82) (323.15, 450.82)      /T1_8 in patient models define mechanisms of cancer immune evasion. 	<|special_separator|>
-(323.15, 433.05) (363.32, 433.05) (363.32, 440.07) (323.15, 440.07)      /T1_9 Nat. Genet.	<|special_separator|>
-(363.18, 433.05) (365.30, 433.05) (365.30, 440.07) (363.18, 440.07)      /T1_8  	<|special_separator|>
-(364.90, 433.07) (374.77, 433.07) (374.77, 440.07) (364.90, 440.07)      /T1_6 53	<|special_separator|>
-(374.63, 433.05) (434.63, 433.05) (434.63, 440.07) (374.63, 440.07)      /T1_8 , 332-341 (2021).	<|special_separator|>
-(306.14, 422.30) (314.66, 422.30) (314.66, 429.31) (306.14, 429.31)      /T1_8 2. 	<|special_separator|>
-(323.15, 422.30) (542.38, 422.30) (542.38, 429.31) (323.15, 429.31)      /T1_8 Dixit, A. et al. Perturb-seq: dissecting molecular circuits with 	<|special_separator|>
-(323.15, 411.55) (543.49, 411.55) (543.49, 418.56) (323.15, 418.56)      /T1_8 scalable single-cell RNA profiling of pooled genetic screens. 	<|special_separator|>
-(543.07, 411.55) (557.53, 411.55) (557.53, 418.56) (543.07, 418.56)      /T1_9 Cell	<|special_separator|>
-(557.65, 411.55) (559.77, 411.55) (559.77, 418.56) (557.65, 418.56)      /T1_8  	<|special_separator|>
-(323.15, 400.81) (335.90, 400.81) (335.90, 407.81) (323.15, 407.81)      /T1_6 167	<|special_separator|>
-(335.76, 400.80) (404.58, 400.80) (404.58, 407.81) (335.76, 407.81)      /T1_8 , 1853-1866 (2016).	<|special_separator|>
-(306.14, 390.04) (314.99, 390.04) (314.99, 397.06) (306.14, 397.06)      /T1_8 3. 	<|special_separator|>
-(323.15, 390.04) (505.68, 390.04) (505.68, 397.06) (323.15, 397.06)      /T1_8 Srivatsan, S. R. et al. Massively multiplex chemical 	<|special_separator|>
-(323.15, 379.29) (471.29, 379.29) (471.29, 386.31) (323.15, 386.31)      /T1_8 transcriptomics at single-cell resolution. 	<|special_separator|>
-(470.87, 379.29) (499.51, 379.29) (499.51, 386.31) (470.87, 386.31)      /T1_9 Science	<|special_separator|>
-(499.37, 379.29) (501.49, 379.29) (501.49, 386.31) (499.37, 386.31)      /T1_8  	<|special_separator|>
-(501.17, 379.31) (515.56, 379.31) (515.56, 386.31) (501.17, 386.31)      /T1_6 367	<|special_separator|>
-(515.41, 379.29) (542.70, 379.29) (542.70, 386.31) (515.41, 386.31)      /T1_8 , 45-51 	<|special_separator|>
-(323.15, 368.54) (349.70, 368.54) (349.70, 375.55) (323.15, 375.55)      /T1_8 (2020).	<|special_separator|>
-(306.14, 357.79) (315.17, 357.79) (315.17, 364.80) (306.14, 364.80)      /T1_8 4. 	<|special_separator|>
-(323.15, 357.79) (550.53, 357.79) (550.53, 364.80) (323.15, 364.80)      /T1_8 Ji, Y ., Lotfollahi, M., Wolf, F. A. & Theis, F. J. Machine learning for 	<|special_separator|>
-(323.15, 347.04) (443.00, 347.04) (443.00, 354.05) (323.15, 354.05)      /T1_8 perturbational single-cell omics. 	<|special_separator|>
-(442.58, 347.04) (475.80, 347.04) (475.80, 354.05) (442.58, 354.05)      /T1_9 Cell Syst.	<|special_separator|>
-(475.66, 347.04) (477.78, 347.04) (477.78, 354.05) (475.66, 354.05)      /T1_8  	<|special_separator|>
-(477.45, 347.05) (485.50, 347.05) (485.50, 354.05) (477.45, 354.05)      /T1_6 12	<|special_separator|>
-(485.37, 347.04) (545.75, 347.04) (545.75, 354.05) (485.37, 354.05)      /T1_8 , 522-537 (2021).	<|special_separator|>
-(306.14, 336.28) (314.69, 336.28) (314.69, 343.30) (306.14, 343.30)      /T1_8 5. 	<|special_separator|>
-(323.15, 336.28) (546.47, 336.28) (546.47, 343.30) (323.15, 343.30)      /T1_8 Peidli, S. et al. scPerturb: harmonized single-cell perturbation 	<|special_separator|>
-(323.15, 325.53) (343.54, 325.53) (343.54, 332.55) (323.15, 332.55)      /T1_8 data. 	<|special_separator|>
-(343.13, 325.53) (390.66, 325.53) (390.66, 332.55) (343.13, 332.55)      /T1_9 Nat. Methods	<|special_separator|>
-(390.51, 325.53) (392.63, 325.53) (392.63, 332.55) (390.51, 332.55)      /T1_8  	<|special_separator|>
-(392.24, 325.55) (400.29, 325.55) (400.29, 332.55) (392.24, 332.55)      /T1_6 21	<|special_separator|>
-(400.16, 325.53) (462.61, 325.53) (462.61, 332.55) (400.16, 332.55)      /T1_8 , 531-540 (2024).	<|special_separator|>
-(306.14, 314.78) (315.06, 314.78) (315.06, 321.79) (306.14, 321.79)      /T1_8 6. 	<|special_separator|>
-(323.15, 314.78) (512.21, 314.78) (512.21, 321.79) (323.15, 321.79)      /T1_8 Ianevski, A. et al. Single-cell transcriptomes identify 	<|special_separator|>
-(323.15, 304.03) (546.89, 304.03) (546.89, 311.04) (323.15, 311.04)      /T1_8 patient-tailored therapies for selective co-inhibition of cancer 	<|special_separator|>
-(323.15, 293.28) (351.74, 293.28) (351.74, 300.29) (323.15, 300.29)      /T1_8 clones. 	<|special_separator|>
-(351.34, 293.28) (403.29, 293.28) (403.29, 300.29) (351.34, 300.29)      /T1_9 Nat. Commun.	<|special_separator|>
-(403.15, 293.28) (405.27, 293.28) (405.27, 300.29) (403.15, 300.29)      /T1_8  	<|special_separator|>
-(404.87, 293.29) (413.10, 293.29) (413.10, 300.29) (404.87, 300.29)      /T1_6 15	<|special_separator|>
-(412.97, 293.28) (462.31, 293.28) (462.31, 300.29) (412.97, 300.29)      /T1_8 , 8579 (2024).	<|special_separator|>
-(306.14, 282.52) (313.71, 282.52) (313.71, 289.54) (306.14, 289.54)      /T1_8 7. 	<|special_separator|>
-(323.15, 282.52) (541.61, 282.52) (541.61, 289.54) (323.15, 289.54)      /T1_8 Bai, Z. et al. Single-cell CAR T atlas reveals type 2 function in 	<|special_separator|>
-(323.15, 271.77) (426.06, 271.77) (426.06, 278.79) (323.15, 278.79)      /T1_8 8-year leukaemia remission. 	<|special_separator|>
-(425.64, 271.77) (449.77, 271.77) (449.77, 278.79) (425.64, 278.79)      /T1_9 Nature	<|special_separator|>
-(449.63, 271.77) (451.75, 271.77) (451.75, 278.79) (449.63, 278.79)      /T1_8  	<|special_separator|>
-(451.38, 271.79) (466.63, 271.79) (466.63, 278.79) (451.38, 278.79)      /T1_6 634	<|special_separator|>
-(466.49, 271.77) (525.24, 271.77) (525.24, 278.79) (466.49, 278.79)      /T1_8 , 702-711 (2024).	<|special_separator|>
-(306.14, 261.02) (314.97, 261.02) (314.97, 268.03) (306.14, 268.03)      /T1_8 8. 	<|special_separator|>
-(323.15, 261.02) (551.60, 261.02) (551.60, 268.03) (323.15, 268.03)      /T1_8 Lotfollahi, M., Wolf, F. A. & Theis, F. J. scGen predicts single-cell 	<|special_separator|>
-(323.15, 250.27) (412.20, 250.27) (412.20, 257.28) (323.15, 257.28)      /T1_8 perturbation responses. 	<|special_separator|>
-(411.78, 250.27) (459.31, 250.27) (459.31, 257.28) (411.78, 257.28)      /T1_9 Nat. Methods	<|special_separator|>
-(459.16, 250.27) (461.28, 250.27) (461.28, 257.28) (459.16, 257.28)      /T1_8  	<|special_separator|>
-(460.93, 250.28) (469.48, 250.28) (469.48, 257.28) (460.93, 257.28)      /T1_6 16	<|special_separator|>
-(469.34, 250.27) (526.52, 250.27) (526.52, 257.28) (469.34, 257.28)      /T1_8 , 715-721 (2019).	<|special_separator|>
-(306.14, 239.52) (314.83, 239.52) (314.83, 246.53) (306.14, 246.53)      /T1_8 9. 	<|special_separator|>
-(323.15, 239.52) (519.89, 239.52) (519.89, 246.53) (323.15, 246.53)      /T1_8 Lotfollahi, M., Naghipourfar, M., Theis, F. J. & Wolf, F. A. 	<|special_separator|>
-(323.15, 228.76) (562.75, 228.76) (562.75, 235.78) (323.15, 235.78)      /T1_8 Conditional out-of-distribution generation for unpaired data using 	<|special_separator|>
-(323.15, 218.01) (371.66, 218.01) (371.66, 225.03) (323.15, 225.03)      /T1_8 transfer VAE. 	<|special_separator|>
-(371.24, 218.01) (422.73, 218.01) (422.73, 225.03) (371.24, 225.03)      /T1_9 Bioinformatics	<|special_separator|>
-(422.59, 218.01) (424.71, 218.01) (424.71, 225.03) (422.59, 225.03)      /T1_8  	<|special_separator|>
-(424.33, 218.03) (434.51, 218.03) (434.51, 225.03) (424.33, 225.03)      /T1_6 36	<|special_separator|>
-(434.37, 218.01) (499.77, 218.01) (499.77, 225.03) (434.37, 225.03)      /T1_8 , i610-i617 (2020).	<|special_separator|>
-(306.14, 207.26) (318.45, 207.26) (318.45, 214.27) (306.14, 214.27)      /T1_8 10. 	<|special_separator|>
-(323.15, 207.26) (541.29, 207.26) (541.29, 214.27) (323.15, 214.27)      /T1_8 Lotfollahi, M. et al. Predicting cellular responses to complex 	<|special_separator|>
-(323.15, 196.51) (477.20, 196.51) (477.20, 203.52) (323.15, 203.52)      /T1_8 perturbations in high-throughput screens. 	<|special_separator|>
-(476.78, 196.51) (528.62, 196.51) (528.62, 203.52) (476.78, 203.52)      /T1_9 Mol. Syst. Biol.	<|special_separator|>
-(528.47, 196.51) (530.59, 196.51) (530.59, 203.52) (528.47, 203.52)      /T1_8  	<|special_separator|>
-(530.28, 196.52) (538.79, 196.52) (538.79, 203.52) (530.28, 203.52)      /T1_6 19	<|special_separator|>
-(538.65, 196.51) (543.03, 196.51) (543.03, 203.52) (538.65, 203.52)      /T1_8 , 	<|special_separator|>
-(323.15, 185.76) (403.60, 185.76) (403.60, 192.77) (323.15, 192.77)      /T1_8 MSB202211517 (2023).	<|special_separator|>
-(306.14, 175.00) (316.43, 175.00) (316.43, 182.02) (306.14, 182.02)      /T1_8 11. 	<|special_separator|>
-(323.15, 175.00) (535.40, 175.00) (535.40, 182.02) (323.15, 182.02)      /T1_8 Hetzel, L. et al. Predicting cellular responses to novel drug 	<|special_separator|>
-(323.15, 164.25) (478.02, 164.25) (478.02, 171.27) (323.15, 171.27)      /T1_8 perturbations at a single-cell resolution. In 	<|special_separator|>
-(477.61, 164.25) (551.09, 164.25) (551.09, 171.27) (477.61, 171.27)      /T1_9 36th Conference on 	<|special_separator|>
-(323.15, 153.50) (517.57, 153.50) (517.57, 160.51) (323.15, 160.51)      /T1_9 Neural Information Processing Systems (NeurIPS 2022)	<|special_separator|>
-(517.43, 153.50) (545.87, 153.50) (545.87, 160.51) (517.43, 160.51)      /T1_8  https://	<|special_separator|>
-(323.15, 142.75) (481.99, 142.75) (481.99, 149.76) (323.15, 149.76)      /T1_8 openreview.net/pdf?id=vRrFVHxFiXJ (2022).	<|special_separator|>
-(306.14, 132.00) (317.75, 132.00) (317.75, 139.01) (306.14, 139.01)      /T1_8 12. 	<|special_separator|>
-(323.15, 132.00) (535.35, 132.00) (535.35, 139.01) (323.15, 139.01)      /T1_8 Yu, H., Qian, W., Song, Y. & Welch, J. D. PerturbNet predicts 	<|special_separator|>
-(323.15, 121.24) (519.86, 121.24) (519.86, 128.26) (323.15, 128.26)      /T1_8 single-cell responses to unseen chemical and genetic 	<|special_separator|>
-(323.15, 110.49) (376.95, 110.49) (376.95, 117.51) (323.15, 117.51)      /T1_8 perturbations. 	<|special_separator|>
-(376.57, 110.49) (428.40, 110.49) (428.40, 117.51) (376.57, 117.51)      /T1_9 Mol. Syst. Biol.	<|special_separator|>
-(428.26, 110.49) (430.38, 110.49) (430.38, 117.51) (428.26, 117.51)      /T1_8  	<|special_separator|>
-(429.98, 110.51) (438.04, 110.51) (438.04, 117.51) (429.98, 117.51)      /T1_6 21	<|special_separator|>
-(437.90, 110.49) (502.16, 110.49) (502.16, 117.51) (437.90, 117.51)      /T1_8 , 960-982 (2025).	<|special_separator|>
-(306.14, 099.74) (318.07, 099.74) (318.07, 106.75) (306.14, 106.75)      /T1_8 13. 	<|special_separator|>
-(323.15, 099.74) (548.97, 099.74) (548.97, 106.75) (323.15, 106.75)      /T1_8 Roohani, Y., Huang, K. & Leskovec, J. Predicting transcriptional 	<|special_separator|>
-(323.15, 088.99) (529.35, 088.99) (529.35, 096.00) (323.15, 096.00)      /T1_8 outcomes of novel multigene perturbations with GEARS. 	<|special_separator|>
-(528.93, 088.99) (545.64, 088.99) (545.64, 096.00) (528.93, 096.00)      /T1_9 Nat. 	<|special_separator|>
-(323.15, 078.24) (363.82, 078.24) (363.82, 085.25) (323.15, 085.25)      /T1_9 Biotechnol.	<|special_separator|>
-(363.68, 078.24) (365.80, 078.24) (365.80, 085.25) (363.68, 085.25)      /T1_8  	<|special_separator|>
-(365.40, 078.25) (375.25, 078.25) (375.25, 085.25) (365.40, 085.25)      /T1_6 42	<|special_separator|>
-(375.12, 078.24) (437.49, 078.24) (437.49, 085.25) (375.12, 085.25)      /T1_8 , 927-935 (2024).	<|special_separator|>
-(306.14, 067.48) (318.10, 067.48) (318.10, 074.50) (306.14, 074.50)      /T1_8 14. 	<|special_separator|>
-(323.15, 067.48) (557.49, 067.48) (557.49, 074.50) (323.15, 074.50)      /T1_8 Zheng, Y. et al. A deep generative model for deciphering cellular 	<|special_separator|>
-(323.15, 056.73) (538.15, 056.73) (538.15, 063.75) (323.15, 063.75)      /T1_8 dynamics and in silico drug discovery in complex diseases. 	<|special_separator|>
-(537.73, 056.73) (554.44, 056.73) (554.44, 063.75) (537.73, 063.75)      /T1_9 Nat. 	<|special_separator|>
-(323.15, 045.98) (370.49, 045.98) (370.49, 052.99) (323.15, 052.99)      /T1_9 Biomed. Eng.	<|special_separator|>
-(370.34, 045.98) (372.46, 045.98) (372.46, 052.99) (370.34, 052.99)      /T1_8  	<|special_separator|>
-(372.07, 046.00) (377.31, 046.00) (377.31, 052.99) (372.07, 052.99)      /T1_6 9	<|special_separator|>
-(377.17, 045.98) (446.52, 045.98) (446.52, 052.99) (377.17, 052.99)      /T1_8 , 2155-2180 (2025).

groundtruth/parser/font_11.pdf.page_no_11.py.json.word.txt

@@ -1,685 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 NatureMachineIntelligence|Volume8|June2026|984-996	<|special_separator|>
-(546.19, 020.07) (562.05, 020.07) (562.05, 027.05) (546.19, 027.05)      /T1_1 994	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 733.65) (294.77, 733.65) (294.77, 741.31) (039.69, 741.31)      /T1_2 networkhas4hiddenlayerswith64neuronseach.Thedoseanddrug	<|special_separator|>
-(039.69, 722.90) (082.66, 722.90) (082.66, 730.56) (039.69, 730.56)      /T1_2 embedders	<|special_separator|>
-(084.26, 722.90) (102.47, 722.90) (102.47, 730.56) (084.26, 730.56)      /T1_2 were	<|special_separator|>
-(104.07, 722.90) (159.66, 722.90) (159.66, 730.56) (104.07, 730.56)      /T1_2 parameterized	<|special_separator|>
-(161.26, 722.90) (177.93, 722.90) (177.93, 730.56) (161.26, 730.56)      /T1_2 with	<|special_separator|>
-(179.53, 722.90) (193.84, 722.90) (193.84, 730.56) (179.53, 730.56)      /T1_2 one	<|special_separator|>
-(195.44, 722.90) (217.89, 722.90) (217.89, 730.56) (195.44, 730.56)      /T1_2 dense	<|special_separator|>
-(219.49, 722.90) (240.02, 722.90) (240.02, 730.56) (219.49, 730.56)      /T1_2 layer.	<|special_separator|>
-(241.62, 722.90) (256.10, 722.90) (256.10, 730.56) (241.62, 730.56)      /T1_2 The	<|special_separator|>
-(257.70, 722.90) (294.73, 722.90) (294.73, 730.56) (257.70, 730.56)      /T1_2 Euclidean	<|special_separator|>
-(039.69, 712.15) (294.82, 712.15) (294.82, 719.81) (039.69, 719.81)      /T1_2 distancewasusedasthedisplacementcost,andtheMongeGapregu-	<|special_separator|>
-(039.69, 701.40) (063.97, 701.40) (063.97, 709.06) (039.69, 709.06)      /T1_2 larizer	<|special_separator|>
-(065.04, 701.41) (069.24, 701.41) (069.24, 709.06) (065.04, 709.06)      /T1_3 λ	<|special_separator|>
-(070.33, 701.40) (085.01, 701.40) (085.01, 709.06) (070.33, 709.06)      /T1_2 =10	<|special_separator|>
-(085.00, 705.59) (090.38, 705.59) (090.38, 710.18) (085.00, 710.18)      /T1_2 -2	<|special_separator|>
-(090.38, 701.40) (294.90, 701.40) (294.90, 709.06) (090.38, 709.06)      /T1_2 .DuringtheOTtrainingphase,werepeatedlysampleda	<|special_separator|>
-(039.69, 690.65) (294.72, 690.65) (294.72, 698.31) (039.69, 698.31)      /T1_2 batchof256observationsfromthesourceandtargetdistributionsfor	<|special_separator|>
-(039.69, 679.90) (294.82, 679.90) (294.82, 687.56) (039.69, 687.56)      /T1_2 1,000iterationsforlocalmodels(withoutaconditionorconditioned	<|special_separator|>
-(039.69, 669.15) (294.71, 669.15) (294.71, 676.81) (039.69, 676.81)      /T1_2 ondose),for10,000iterationsfortheglobalmodels(RDKitandMoA)	<|special_separator|>
-(039.69, 658.40) (054.21, 658.40) (054.21, 666.06) (039.69, 666.06)      /T1_2 and	<|special_separator|>
-(055.90, 658.40) (067.51, 658.40) (067.51, 666.06) (055.90, 666.06)      /T1_2 for	<|special_separator|>
-(069.21, 658.40) (103.32, 658.40) (103.32, 666.06) (069.21, 666.06)      /T1_2 500,000	<|special_separator|>
-(105.02, 658.40) (142.44, 658.40) (142.44, 666.06) (105.02, 666.06)      /T1_2 iterations	<|special_separator|>
-(144.13, 658.40) (155.74, 658.40) (155.74, 666.06) (144.13, 666.06)      /T1_2 for	<|special_separator|>
-(157.44, 658.40) (170.15, 658.40) (170.15, 666.06) (157.44, 666.06)      /T1_2 the	<|special_separator|>
-(171.84, 658.40) (196.96, 658.40) (196.96, 666.06) (171.84, 666.06)      /T1_2 bigger	<|special_separator|>
-(198.65, 658.40) (227.20, 658.40) (227.20, 666.06) (198.65, 666.06)      /T1_2 models	<|special_separator|>
-(228.89, 658.40) (244.73, 658.40) (244.73, 666.06) (228.89, 666.06)      /T1_2 that	<|special_separator|>
-(246.42, 658.40) (264.96, 658.40) (264.96, 666.06) (246.42, 666.06)      /T1_2 were	<|special_separator|>
-(266.66, 658.40) (294.75, 658.40) (294.75, 666.06) (266.66, 666.06)      /T1_2 trained	<|special_separator|>
-(039.69, 647.65) (049.90, 647.65) (049.90, 655.31) (039.69, 655.31)      /T1_2 on	<|special_separator|>
-(051.85, 647.65) (061.21, 647.65) (061.21, 655.31) (051.85, 655.31)      /T1_2 all	<|special_separator|>
-(063.16, 647.65) (085.94, 647.65) (085.94, 655.31) (063.16, 655.31)      /T1_2 drugs	<|special_separator|>
-(087.89, 647.65) (095.42, 647.65) (095.42, 655.31) (087.89, 655.31)      /T1_2 in	<|special_separator|>
-(097.38, 647.65) (110.31, 647.65) (110.31, 655.31) (097.38, 655.31)      /T1_2 the	<|special_separator|>
-(112.26, 647.65) (140.53, 647.65) (140.53, 655.31) (112.26, 655.31)      /T1_2 SciPlex	<|special_separator|>
-(142.48, 647.65) (174.00, 647.65) (174.00, 655.31) (142.48, 655.31)      /T1_2 dataset.	<|special_separator|>
-(175.95, 647.65) (194.60, 647.65) (194.60, 655.31) (175.95, 655.31)      /T1_2 Each	<|special_separator|>
-(196.55, 647.65) (218.82, 647.65) (218.82, 655.31) (196.55, 655.31)      /T1_2 batch	<|special_separator|>
-(220.77, 647.65) (260.62, 647.65) (260.62, 655.31) (220.77, 655.31)      /T1_2 contained	<|special_separator|>
-(262.57, 647.65) (294.85, 647.65) (294.85, 655.31) (262.57, 655.31)      /T1_2 samples	<|special_separator|>
-(039.69, 636.90) (294.79, 636.90) (294.79, 644.56) (039.69, 644.56)      /T1_2 fromonlyonecontext,whichwasuniformlysampled.FortheSciPlex	<|special_separator|>
-(039.69, 626.15) (090.52, 626.15) (090.52, 633.81) (039.69, 633.81)      /T1_2 experiments,	<|special_separator|>
-(092.19, 626.15) (104.86, 626.15) (104.86, 633.81) (092.19, 633.81)      /T1_2 the	<|special_separator|>
-(106.53, 626.15) (156.03, 626.15) (156.03, 633.81) (106.53, 633.81)      /T1_2 autoencoder	<|special_separator|>
-(157.70, 626.15) (194.75, 626.15) (194.75, 633.81) (157.70, 633.81)      /T1_2 consisted	<|special_separator|>
-(196.42, 626.15) (204.46, 626.15) (204.46, 633.81) (196.42, 633.81)      /T1_2 of	<|special_separator|>
-(206.13, 626.15) (220.61, 626.15) (220.61, 633.81) (206.13, 633.81)      /T1_2 two	<|special_separator|>
-(222.29, 626.15) (249.46, 626.15) (249.46, 633.81) (222.29, 633.81)      /T1_2 hidden	<|special_separator|>
-(251.13, 626.15) (276.15, 626.15) (276.15, 633.81) (251.13, 633.81)      /T1_2 layers,	<|special_separator|>
-(277.82, 626.15) (294.74, 626.15) (294.74, 633.81) (277.82, 633.81)      /T1_2 with	<|special_separator|>
-(039.69, 615.40) (294.78, 615.40) (294.78, 623.06) (039.69, 623.06)      /T1_2 512dimensionsinboththeencoderanddecoder.The50-dimensional	<|special_separator|>
-(039.69, 604.65) (294.80, 604.65) (294.80, 612.31) (039.69, 612.31)      /T1_2 latentrepresentationwaslearnedthrough50epochswithabatchsize	<|special_separator|>
-(039.69, 593.90) (212.65, 593.90) (212.65, 601.56) (039.69, 601.56)      /T1_2 of256.Allmodelswereimplementedusingthe	<|special_separator|>
-(213.89, 594.22) (247.80, 594.22) (247.80, 600.33) (213.89, 600.33)      /T1_4 OTT-JAX	<|special_separator|>
-(248.91, 593.90) (280.14, 593.90) (280.14, 601.56) (248.91, 601.56)      /T1_2 package	<|special_separator|>
-(280.11, 598.09) (286.03, 598.09) (286.03, 602.68) (280.11, 602.68)      /T1_2 64	<|special_separator|>
-(285.96, 593.90) (288.18, 593.90) (288.18, 601.56) (285.96, 601.56)      /T1_2 .	<|special_separator|>
-(056.69, 583.15) (069.37, 583.15) (069.37, 590.81) (056.69, 590.81)      /T1_2 For	<|special_separator|>
-(070.79, 583.15) (083.27, 583.15) (083.27, 590.81) (070.79, 590.81)      /T1_2 the	<|special_separator|>
-(084.69, 583.15) (125.12, 583.15) (125.12, 590.81) (084.69, 590.81)      /T1_2 large-scale	<|special_separator|>
-(126.53, 583.15) (153.65, 583.15) (153.65, 590.81) (126.53, 590.81)      /T1_2 SciPlex	<|special_separator|>
-(155.07, 583.15) (247.36, 583.15) (247.36, 590.81) (155.07, 590.81)      /T1_2 experiment(712training	<|special_separator|>
-(248.77, 583.15) (294.78, 583.15) (294.78, 590.81) (248.77, 590.81)      /T1_2 conditions),	<|special_separator|>
-(039.69, 572.40) (294.81, 572.40) (294.81, 580.06) (039.69, 580.06)      /T1_2 weincreasedthemodelsizeofCMongeandtheembeddingsizeofthe	<|special_separator|>
-(039.69, 561.65) (058.03, 561.65) (058.03, 569.31) (039.69, 569.31)      /T1_2 gene	<|special_separator|>
-(059.63, 561.65) (100.87, 561.65) (100.87, 569.31) (059.63, 569.31)      /T1_2 expression	<|special_separator|>
-(102.47, 561.65) (116.74, 561.65) (116.74, 569.31) (102.47, 569.31)      /T1_2 and	<|special_separator|>
-(118.35, 561.65) (130.83, 561.65) (130.83, 569.31) (118.35, 569.31)      /T1_2 the	<|special_separator|>
-(132.43, 561.65) (163.73, 561.65) (163.73, 569.31) (132.43, 569.31)      /T1_2 context.	<|special_separator|>
-(165.33, 561.65) (179.82, 561.65) (179.82, 569.31) (165.33, 569.31)      /T1_2 The	<|special_separator|>
-(181.42, 561.65) (199.66, 561.65) (199.66, 569.31) (181.42, 569.31)      /T1_2 drug	<|special_separator|>
-(201.26, 561.65) (215.54, 561.65) (215.54, 569.31) (201.26, 569.31)      /T1_2 and	<|special_separator|>
-(217.14, 561.65) (233.92, 561.65) (233.92, 569.31) (217.14, 569.31)      /T1_2 data	<|special_separator|>
-(235.52, 561.65) (279.02, 561.65) (279.02, 569.31) (235.52, 569.31)      /T1_2 embedding	<|special_separator|>
-(280.62, 561.65) (294.80, 561.65) (294.80, 569.31) (280.62, 569.31)      /T1_2 was	<|special_separator|>
-(039.69, 550.90) (294.83, 550.90) (294.83, 558.56) (039.69, 558.56)      /T1_2 increasedfrom50to100dimensions,whichincreasedthenumberof	<|special_separator|>
-(039.69, 540.15) (083.07, 540.15) (083.07, 547.81) (039.69, 547.81)      /T1_2 parameters	<|special_separator|>
-(084.65, 540.15) (091.93, 540.15) (091.93, 547.81) (084.65, 547.81)      /T1_2 in	<|special_separator|>
-(093.50, 540.15) (105.99, 540.15) (105.99, 547.81) (093.50, 547.81)      /T1_2 the	<|special_separator|>
-(107.56, 540.15) (125.90, 540.15) (125.90, 547.81) (107.56, 547.81)      /T1_2 gene	<|special_separator|>
-(127.47, 540.15) (168.69, 540.15) (168.69, 547.81) (127.47, 547.81)      /T1_2 expression	<|special_separator|>
-(170.27, 540.15) (218.97, 540.15) (218.97, 547.81) (170.27, 547.81)      /T1_2 autoencoder	<|special_separator|>
-(220.54, 540.15) (239.27, 540.15) (239.27, 547.81) (220.54, 547.81)      /T1_2 from	<|special_separator|>
-(240.84, 540.15) (256.75, 540.15) (256.75, 547.81) (240.84, 547.81)      /T1_2 1.60	<|special_separator|>
-(258.32, 540.15) (285.13, 540.15) (285.13, 547.81) (258.32, 547.81)      /T1_2 million	<|special_separator|>
-(286.70, 540.15) (294.81, 540.15) (294.81, 547.81) (286.70, 547.81)      /T1_2 to	<|special_separator|>
-(039.69, 529.40) (294.77, 529.40) (294.77, 537.06) (039.69, 537.06)      /T1_2 1.65million.ThefourfullyconnectedlayersofCMongewereincreased	<|special_separator|>
-(039.69, 518.65) (058.66, 518.65) (058.66, 526.31) (039.69, 526.31)      /T1_2 from	<|special_separator|>
-(060.33, 518.65) (075.53, 518.65) (075.53, 526.31) (060.33, 526.31)      /T1_2 [64,	<|special_separator|>
-(077.20, 518.65) (089.37, 518.65) (089.37, 526.31) (077.20, 526.31)      /T1_2 64,	<|special_separator|>
-(091.04, 518.65) (103.22, 518.65) (103.22, 526.31) (091.04, 526.31)      /T1_2 64,	<|special_separator|>
-(104.89, 518.65) (117.90, 518.65) (117.90, 526.31) (104.89, 526.31)      /T1_2 64]	<|special_separator|>
-(119.56, 518.65) (127.77, 518.65) (127.77, 526.31) (119.56, 526.31)      /T1_2 to	<|special_separator|>
-(129.44, 518.65) (148.40, 518.65) (148.40, 526.31) (129.44, 526.31)      /T1_2 [256,	<|special_separator|>
-(150.07, 518.65) (166.09, 518.65) (166.09, 526.31) (150.07, 526.31)      /T1_2 256,	<|special_separator|>
-(167.76, 518.65) (182.46, 518.65) (182.46, 526.31) (167.76, 526.31)      /T1_2 512,	<|special_separator|>
-(184.13, 518.65) (201.85, 518.65) (201.85, 526.31) (184.13, 526.31)      /T1_2 512],	<|special_separator|>
-(203.52, 518.65) (237.96, 518.65) (237.96, 526.31) (203.52, 526.31)      /T1_2 resulting	<|special_separator|>
-(239.63, 518.65) (247.00, 518.65) (247.00, 526.31) (239.63, 526.31)      /T1_2 in	<|special_separator|>
-(248.67, 518.65) (252.98, 518.65) (252.98, 526.31) (248.67, 526.31)      /T1_2 a	<|special_separator|>
-(254.65, 518.65) (294.85, 518.65) (294.85, 526.31) (254.65, 526.31)      /T1_2 parameter	<|special_separator|>
-(039.69, 507.90) (072.54, 507.90) (072.54, 515.56) (039.69, 515.56)      /T1_2 increase	<|special_separator|>
-(074.52, 507.90) (093.94, 507.90) (093.94, 515.56) (074.52, 515.56)      /T1_2 from	<|special_separator|>
-(095.91, 507.90) (124.43, 507.90) (124.43, 515.56) (095.91, 515.56)      /T1_2 23,000	<|special_separator|>
-(126.41, 507.90) (134.78, 507.90) (134.78, 515.56) (126.41, 515.56)      /T1_2 to	<|special_separator|>
-(136.76, 507.90) (170.96, 507.90) (170.96, 515.56) (136.76, 515.56)      /T1_2 560,000	<|special_separator|>
-(172.94, 507.90) (213.34, 507.90) (213.34, 515.56) (172.94, 515.56)      /T1_2 (580,000)	<|special_separator|>
-(215.32, 507.90) (227.14, 507.90) (227.14, 515.56) (215.32, 515.56)      /T1_2 for	<|special_separator|>
-(229.12, 507.90) (294.70, 507.90) (294.70, 515.56) (229.12, 515.56)      /T1_2 MoAembedding	<|special_separator|>
-(039.69, 497.15) (065.09, 497.15) (065.09, 504.81) (039.69, 504.81)      /T1_2 (RDKit	<|special_separator|>
-(066.72, 497.15) (112.79, 497.15) (112.79, 504.81) (066.72, 504.81)      /T1_2 embedding,	<|special_separator|>
-(114.42, 497.15) (165.95, 497.15) (165.95, 504.81) (114.42, 504.81)      /T1_2 respectively).	<|special_separator|>
-(167.58, 497.15) (180.34, 497.15) (180.34, 504.81) (167.58, 504.81)      /T1_2 For	<|special_separator|>
-(181.97, 497.15) (229.95, 497.15) (229.95, 504.81) (181.97, 504.81)      /T1_2 comparison,	<|special_separator|>
-(231.58, 497.15) (270.51, 497.15) (270.51, 504.81) (231.58, 504.81)      /T1_2 chemCPA,	<|special_separator|>
-(272.14, 497.15) (294.94, 497.15) (294.94, 504.81) (272.14, 504.81)      /T1_2 which	<|special_separator|>
-(039.69, 486.40) (294.80, 486.40) (294.80, 494.06) (039.69, 494.06)      /T1_2 hasonemodelforembeddingthedata,contextinformationandlearn-	<|special_separator|>
-(039.69, 475.65) (251.63, 475.65) (251.63, 483.31) (039.69, 483.31)      /T1_2 ingtheperturbation,hasaround1.37millionparameters.	<|special_separator|>
-(039.69, 454.19) (100.40, 454.19) (100.40, 462.17) (039.69, 462.17)      /T1_5 Benchmarking	<|special_separator|>
-(039.69, 443.40) (294.80, 443.40) (294.80, 451.06) (039.69, 451.06)      /T1_2 Wefirstbenchmarkeddifferentstate-of-the-artmethodsforuncondi-	<|special_separator|>
-(039.69, 432.65) (154.31, 432.65) (154.31, 440.31) (039.69, 440.31)      /T1_2 tionalperturbationmodelling.	<|special_separator|>
-(039.69, 411.53) (047.66, 411.53) (047.66, 418.75) (039.69, 418.75)      /T1_6 4i	<|special_separator|>
-(049.92, 411.53) (080.88, 411.53) (080.88, 418.75) (049.92, 418.75)      /T1_6 dataset	<|special_separator|>
-(080.97, 411.29) (083.46, 411.29) (083.46, 418.81) (080.97, 418.81)      /T1_5 .	<|special_separator|>
-(085.27, 411.15) (098.36, 411.15) (098.36, 418.81) (085.27, 418.81)      /T1_2 For	<|special_separator|>
-(100.15, 411.15) (113.03, 411.15) (113.03, 418.81) (100.15, 418.81)      /T1_2 the	<|special_separator|>
-(114.82, 411.15) (178.12, 411.15) (178.12, 418.81) (114.82, 418.81)      /T1_2 non-conditional	<|special_separator|>
-(179.90, 411.15) (231.72, 411.15) (231.72, 418.81) (179.90, 418.81)      /T1_2 experiments,	<|special_separator|>
-(233.51, 411.15) (244.31, 411.15) (244.31, 418.81) (233.51, 418.81)      /T1_2 we	<|special_separator|>
-(246.10, 411.15) (274.62, 411.15) (274.62, 418.81) (246.10, 418.81)      /T1_2 trained	<|special_separator|>
-(276.41, 411.15) (294.81, 411.15) (294.81, 418.81) (276.41, 418.81)      /T1_2 each	<|special_separator|>
-(039.69, 400.40) (294.78, 400.40) (294.78, 408.06) (039.69, 408.06)      /T1_2 methodoneachofthe35therapieswithan80/20training/validation	<|special_separator|>
-(039.69, 389.65) (176.62, 389.65) (176.62, 397.31) (039.69, 397.31)      /T1_2 split.WenotethattheoriginalscGen	<|special_separator|>
-(176.60, 393.84) (179.52, 393.84) (179.52, 398.43) (176.60, 398.43)      /T1_2 8	<|special_separator|>
-(180.90, 389.65) (201.43, 389.65) (201.43, 397.31) (180.90, 397.31)      /T1_2 relies	<|special_separator|>
-(202.81, 389.65) (260.47, 389.65) (260.47, 397.31) (202.81, 397.31)      /T1_2 onavariational	<|special_separator|>
-(261.85, 389.65) (294.82, 389.65) (294.82, 397.31) (261.85, 397.31)      /T1_2 formula-	<|special_separator|>
-(039.69, 378.90) (055.31, 378.90) (055.31, 386.56) (039.69, 386.56)      /T1_2 tion	<|special_separator|>
-(055.35, 383.09) (060.85, 383.09) (060.85, 387.68) (055.35, 387.68)      /T1_2 65	<|special_separator|>
-(060.74, 378.90) (062.99, 378.90) (062.99, 386.56) (060.74, 386.56)      /T1_2 .	<|special_separator|>
-(064.65, 378.90) (095.28, 378.90) (095.28, 386.56) (064.65, 386.56)      /T1_2 Instead,	<|special_separator|>
-(096.94, 378.90) (104.31, 378.90) (104.31, 386.56) (096.94, 386.56)      /T1_2 in	<|special_separator|>
-(105.97, 378.90) (119.48, 378.90) (119.48, 386.56) (105.97, 386.56)      /T1_2 our	<|special_separator|>
-(121.14, 378.90) (171.79, 378.90) (171.79, 386.56) (121.14, 386.56)      /T1_2 experiments,	<|special_separator|>
-(173.46, 378.90) (219.30, 378.90) (219.30, 386.56) (173.46, 386.56)      /T1_2 wefollowed	<|special_separator|>
-(220.96, 378.90) (233.59, 378.90) (233.59, 386.56) (220.96, 386.56)      /T1_2 the	<|special_separator|>
-(235.25, 378.90) (259.35, 378.90) (259.35, 386.56) (235.25, 386.56)      /T1_2 set-up	<|special_separator|>
-(261.01, 378.90) (268.37, 378.90) (268.37, 386.56) (261.01, 386.56)      /T1_2 in	<|special_separator|>
-(270.04, 378.90) (294.71, 378.90) (294.71, 386.56) (270.04, 386.56)      /T1_2 Bunne	<|special_separator|>
-(039.69, 368.15) (047.27, 368.15) (047.27, 375.81) (039.69, 375.81)      /T1_2 et	<|special_separator|>
-(048.89, 368.15) (057.79, 368.15) (057.79, 375.81) (048.89, 375.81)      /T1_2 al.	<|special_separator|>
-(057.81, 372.34) (063.17, 372.34) (063.17, 376.93) (057.81, 376.93)      /T1_2 33	<|special_separator|>
-(063.11, 368.15) (065.34, 368.15) (065.34, 375.81) (063.11, 375.81)      /T1_2 .	<|special_separator|>
-(066.96, 368.15) (084.57, 368.15) (084.57, 375.81) (066.96, 375.81)      /T1_2 That	<|special_separator|>
-(086.19, 368.15) (094.60, 368.15) (094.60, 375.81) (086.19, 375.81)      /T1_2 is,	<|special_separator|>
-(096.22, 368.15) (159.04, 368.15) (159.04, 375.81) (096.22, 375.81)      /T1_2 weusedavanilla	<|special_separator|>
-(160.66, 368.15) (211.61, 368.15) (211.61, 375.81) (160.66, 375.81)      /T1_2 autoencoder,	<|special_separator|>
-(213.23, 368.15) (227.56, 368.15) (227.56, 375.81) (213.23, 375.81)      /T1_2 and	<|special_separator|>
-(229.19, 368.15) (247.36, 368.15) (247.36, 375.81) (229.19, 375.81)      /T1_2 both	<|special_separator|>
-(248.98, 368.15) (261.52, 368.15) (261.52, 375.81) (248.98, 375.81)      /T1_2 the	<|special_separator|>
-(263.15, 368.15) (294.67, 368.15) (294.67, 375.81) (263.15, 375.81)      /T1_2 encoder	<|special_separator|>
-(039.69, 357.40) (294.81, 357.40) (294.81, 365.06) (039.69, 365.06)      /T1_2 andthedecoderwereparameterizedwithfullyconnectedlayers.The	<|special_separator|>
-(039.69, 346.65) (065.94, 346.65) (065.94, 354.31) (039.69, 354.31)      /T1_2 results	<|special_separator|>
-(067.66, 346.65) (075.10, 346.65) (075.10, 354.31) (067.66, 354.31)      /T1_2 in	<|special_separator|>
-(076.81, 346.65) (136.87, 346.65) (136.87, 354.31) (076.81, 354.31)      /T1_2 Supplementary	<|special_separator|>
-(138.59, 346.65) (159.77, 346.65) (159.77, 354.31) (138.59, 354.31)      /T1_2 Table	<|special_separator|>
-(161.49, 346.65) (166.03, 346.65) (166.03, 354.31) (161.49, 354.31)      /T1_2 2	<|special_separator|>
-(167.74, 346.65) (182.32, 346.65) (182.32, 354.31) (167.74, 354.31)      /T1_2 and	<|special_separator|>
-(184.04, 346.65) (220.87, 346.65) (220.87, 354.31) (184.04, 354.31)      /T1_2 Extended	<|special_separator|>
-(222.58, 346.65) (240.60, 346.65) (240.60, 354.31) (222.58, 354.31)      /T1_2 Data	<|special_separator|>
-(242.32, 346.65) (256.58, 346.65) (256.58, 354.31) (242.32, 354.31)      /T1_2 Fig.	<|special_separator|>
-(258.29, 346.65) (261.79, 346.65) (261.79, 354.31) (258.29, 354.31)      /T1_2 1	<|special_separator|>
-(263.51, 346.65) (294.67, 346.65) (294.67, 354.31) (263.51, 354.31)      /T1_2 confirm	<|special_separator|>
-(039.69, 335.90) (052.54, 335.90) (052.54, 343.56) (039.69, 343.56)      /T1_2 the	<|special_separator|>
-(054.29, 335.90) (082.72, 335.90) (082.72, 343.56) (054.29, 343.56)      /T1_2 finding	<|special_separator|>
-(084.48, 335.90) (093.99, 335.90) (093.99, 343.56) (084.48, 343.56)      /T1_2 by	<|special_separator|>
-(095.74, 335.90) (108.79, 335.90) (108.79, 343.56) (095.74, 343.56)      /T1_2 ref.	<|special_separator|>
-(110.55, 335.90) (122.03, 335.90) (122.03, 343.56) (110.55, 343.56)      /T1_2 32,	<|special_separator|>
-(123.79, 335.90) (152.39, 335.90) (152.39, 343.56) (123.79, 343.56)      /T1_2 namely	<|special_separator|>
-(154.15, 335.90) (170.17, 335.90) (170.17, 343.56) (154.15, 343.56)      /T1_2 that	<|special_separator|>
-(171.93, 335.90) (251.77, 335.90) (251.77, 343.56) (171.93, 343.56)      /T1_2 MongeGapachieves	<|special_separator|>
-(253.52, 335.90) (280.16, 335.90) (280.16, 343.56) (253.52, 343.56)      /T1_2 overall	<|special_separator|>
-(281.92, 335.90) (294.77, 335.90) (294.77, 343.56) (281.92, 343.56)      /T1_2 the	<|special_separator|>
-(039.69, 325.15) (294.82, 325.15) (294.82, 332.81) (039.69, 332.81)      /T1_2 bestresultwithrespecttoeachoftheevaluationmetricsandalsohas	<|special_separator|>
-(039.69, 314.40) (044.02, 314.40) (044.02, 322.06) (039.69, 322.06)      /T1_2 a	<|special_separator|>
-(045.73, 314.40) (067.37, 314.40) (067.37, 322.06) (045.73, 322.06)      /T1_2 lower	<|special_separator|>
-(069.08, 314.40) (103.66, 314.40) (103.66, 322.06) (069.08, 322.06)      /T1_2 standard	<|special_separator|>
-(105.37, 314.40) (144.39, 314.40) (144.39, 322.06) (105.37, 322.06)      /T1_2 deviation.	<|special_separator|>
-(146.10, 314.40) (153.69, 314.40) (153.69, 322.06) (146.10, 322.06)      /T1_2 In	<|special_separator|>
-(155.41, 314.40) (192.19, 314.40) (192.19, 322.06) (155.41, 322.06)      /T1_2 Extended	<|special_separator|>
-(193.90, 314.40) (211.90, 314.40) (211.90, 322.06) (193.90, 322.06)      /T1_2 Data	<|special_separator|>
-(213.61, 314.40) (227.85, 314.40) (227.85, 322.06) (213.61, 322.06)      /T1_2 Fig.	<|special_separator|>
-(229.57, 314.40) (235.41, 314.40) (235.41, 322.06) (229.57, 322.06)      /T1_2 1,	<|special_separator|>
-(237.12, 314.40) (249.87, 314.40) (249.87, 322.06) (237.12, 322.06)      /T1_2 the	<|special_separator|>
-(251.59, 314.40) (281.53, 314.40) (281.53, 322.06) (251.59, 322.06)      /T1_2 subplot	<|special_separator|>
-(283.24, 314.40) (294.89, 314.40) (294.89, 322.06) (283.24, 322.06)      /T1_2 for	<|special_separator|>
-(039.69, 303.65) (294.80, 303.65) (294.80, 311.31) (039.69, 311.31)      /T1_2 theWassersteindistanceclearlyshowsthattheMongemodel(which	<|special_separator|>
-(039.69, 292.90) (068.92, 292.90) (068.92, 300.56) (039.69, 300.56)      /T1_2 directly	<|special_separator|>
-(070.29, 292.90) (123.50, 292.90) (123.50, 300.56) (070.29, 300.56)      /T1_2 optimizesthis	<|special_separator|>
-(124.87, 292.90) (152.88, 292.90) (152.88, 300.56) (124.87, 300.56)      /T1_2 metric)	<|special_separator|>
-(154.25, 292.90) (244.93, 292.90) (244.93, 300.56) (154.25, 300.56)      /T1_2 performedconsistently,	<|special_separator|>
-(246.30, 292.90) (285.48, 292.90) (285.48, 300.56) (246.30, 300.56)      /T1_2 regardless	<|special_separator|>
-(286.85, 292.90) (294.78, 292.90) (294.78, 300.56) (286.85, 300.56)      /T1_2 of	<|special_separator|>
-(039.69, 282.15) (052.17, 282.15) (052.17, 289.81) (039.69, 289.81)      /T1_2 the	<|special_separator|>
-(053.69, 282.15) (105.12, 282.15) (105.12, 289.81) (053.69, 289.81)      /T1_2 perturbation,	<|special_separator|>
-(106.64, 282.15) (137.87, 282.15) (137.87, 289.81) (106.64, 289.81)      /T1_2 whereas	<|special_separator|>
-(139.39, 282.15) (151.88, 282.15) (151.88, 289.81) (139.39, 289.81)      /T1_2 the	<|special_separator|>
-(153.40, 282.15) (202.10, 282.15) (202.10, 289.81) (153.40, 289.81)      /T1_2 autoencoder	<|special_separator|>
-(203.63, 282.15) (240.27, 282.15) (240.27, 289.81) (203.63, 289.81)      /T1_2 struggled	<|special_separator|>
-(241.79, 282.15) (249.90, 282.15) (249.90, 289.81) (241.79, 289.81)      /T1_2 to	<|special_separator|>
-(251.43, 282.15) (280.82, 282.15) (280.82, 289.81) (251.43, 289.81)      /T1_2 capture	<|special_separator|>
-(282.34, 282.15) (294.82, 282.15) (294.82, 289.81) (282.34, 289.81)      /T1_2 the	<|special_separator|>
-(039.69, 271.40) (294.90, 271.40) (294.90, 279.06) (039.69, 279.06)      /T1_2 effectsoftheperturbationinsomecases.Ontheotherhand,theICNN	<|special_separator|>
-(039.69, 260.65) (065.45, 260.65) (065.45, 268.31) (039.69, 268.31)      /T1_2 results	<|special_separator|>
-(067.08, 260.65) (079.12, 260.65) (079.12, 268.31) (067.08, 268.31)      /T1_2 are	<|special_separator|>
-(080.75, 260.65) (108.94, 260.65) (108.94, 268.31) (080.75, 268.31)      /T1_2 skewed	<|special_separator|>
-(110.57, 260.65) (119.88, 260.65) (119.88, 268.31) (110.57, 268.31)      /T1_2 by	<|special_separator|>
-(121.51, 260.65) (125.80, 260.65) (125.80, 268.31) (121.51, 268.31)      /T1_2 a	<|special_separator|>
-(127.43, 260.65) (150.17, 260.65) (150.17, 268.31) (127.43, 268.31)      /T1_2 single	<|special_separator|>
-(151.80, 260.65) (179.57, 260.65) (179.57, 268.31) (151.80, 268.31)      /T1_2 outlier.	<|special_separator|>
-(181.20, 260.65) (213.48, 260.65) (213.48, 268.31) (181.20, 268.31)      /T1_2 Notably,	<|special_separator|>
-(215.11, 260.65) (227.67, 260.65) (227.67, 268.31) (215.11, 268.31)      /T1_2 the	<|special_separator|>
-(229.30, 260.65) (265.09, 260.65) (265.09, 268.31) (229.30, 268.31)      /T1_2 OT-based	<|special_separator|>
-(266.73, 260.65) (294.92, 260.65) (294.92, 268.31) (266.73, 268.31)      /T1_2 models	<|special_separator|>
-(039.69, 249.90) (190.56, 249.90) (190.56, 257.56) (039.69, 257.56)      /T1_2 outperformedtheautoencoderonMMD,	<|special_separator|>
-(191.61, 249.91) (196.72, 249.91) (196.72, 257.56) (191.61, 257.56)      /T1_3 R	<|special_separator|>
-(196.70, 254.09) (199.36, 254.09) (199.36, 258.68) (196.70, 258.68)      /T1_2 2	<|special_separator|>
-(200.41, 249.90) (294.81, 249.90) (294.81, 257.56) (200.41, 257.56)      /T1_2 anddrugsignatures,even	<|special_separator|>
-(039.69, 239.15) (282.05, 239.15) (282.05, 246.81) (039.69, 246.81)      /T1_2 thoughtheyweretrainedtooptimizetheprimalanddualOTloss.	<|special_separator|>
-(039.69, 218.03) (068.61, 218.03) (068.61, 225.25) (039.69, 225.25)      /T1_6 SciPlex	<|special_separator|>
-(068.56, 217.79) (070.97, 217.79) (070.97, 225.31) (068.56, 225.31)      /T1_5 .	<|special_separator|>
-(071.99, 217.65) (294.79, 217.65) (294.79, 225.31) (071.99, 225.31)      /T1_2 Foreachoftheninedrugsandeachofthefourdoses,wefitted	<|special_separator|>
-(039.69, 206.90) (294.77, 206.90) (294.77, 214.56) (039.69, 214.56)      /T1_2 adifferentmodel,resultingin36modelspermethod.Weincludedthe	<|special_separator|>
-(039.69, 196.15) (294.93, 196.15) (294.93, 203.81) (039.69, 203.81)      /T1_2 identitymappingasabaseline,whichsimplypredictstheunperturbed	<|special_separator|>
-(039.69, 185.40) (294.87, 185.40) (294.87, 193.06) (039.69, 193.06)      /T1_2 cellstates.SupplementaryTable1andSupplementaryFig.4showthe	<|special_separator|>
-(039.69, 174.65) (090.02, 174.65) (090.02, 182.31) (039.69, 182.31)      /T1_2 performance	<|special_separator|>
-(091.70, 174.65) (099.73, 174.65) (099.73, 182.31) (091.70, 182.31)      /T1_2 of	<|special_separator|>
-(101.41, 174.65) (114.07, 174.65) (114.07, 182.31) (101.41, 182.31)      /T1_2 the	<|special_separator|>
-(115.75, 174.65) (149.67, 174.65) (149.67, 182.31) (115.75, 182.31)      /T1_2 different	<|special_separator|>
-(151.35, 174.65) (187.89, 174.65) (187.89, 182.31) (151.35, 182.31)      /T1_2 methods.	<|special_separator|>
-(189.57, 174.65) (204.46, 174.65) (204.46, 182.31) (189.57, 182.31)      /T1_2 Our	<|special_separator|>
-(206.13, 174.65) (225.43, 174.65) (225.43, 182.31) (206.13, 182.31)      /T1_2 main	<|special_separator|>
-(227.11, 174.65) (267.66, 174.65) (267.66, 182.31) (227.11, 182.31)      /T1_2 evaluation	<|special_separator|>
-(269.33, 174.65) (294.76, 174.65) (294.76, 182.31) (269.33, 182.31)      /T1_2 metric	<|special_separator|>
-(039.69, 163.90) (294.80, 163.90) (294.80, 171.56) (039.69, 171.56)      /T1_2 wasstillthecoefficientofdeterminationofthefeaturemeansforthe	<|special_separator|>
-(039.69, 153.15) (122.53, 153.15) (122.53, 160.81) (039.69, 160.81)      /T1_2 highlyvariablegenes(	<|special_separator|>
-(122.53, 153.16) (127.68, 153.16) (127.68, 160.81) (122.53, 160.81)      /T1_3 R	<|special_separator|>
-(127.68, 157.34) (130.36, 157.34) (130.36, 161.93) (127.68, 161.93)      /T1_2 2	<|special_separator|>
-(130.36, 153.15) (294.81, 153.15) (294.81, 160.81) (130.36, 160.81)      /T1_2 ),butwealsoreporttheentropy-regularized	<|special_separator|>
-(039.69, 142.40) (294.75, 142.40) (294.75, 150.06) (039.69, 150.06)      /T1_2 Wassersteindistance,whichiscloselyrelatedtotheobjectivefunction	<|special_separator|>
-(039.69, 131.65) (294.86, 131.65) (294.86, 139.31) (039.69, 139.31)      /T1_2 oftheICNNandMongemodels.TheICNNandMongemodelsusedthe	<|special_separator|>
-(039.69, 120.90) (285.89, 120.90) (285.89, 128.56) (039.69, 128.56)      /T1_2 50-dimensionallatentrepresentationlearnedbytheautoencoder.	<|special_separator|>
-(056.69, 110.15) (294.73, 110.15) (294.73, 117.81) (056.69, 117.81)      /T1_2 Allmodelpredictionsweredecodedandevaluatedinthecellspace	<|special_separator|>
-(039.69, 099.40) (294.92, 099.40) (294.92, 107.06) (039.69, 107.06)      /T1_2 forthe50highlyvariablegenes.Weobservedthattheneural-OT-based	<|special_separator|>
-(039.69, 088.65) (273.44, 088.65) (273.44, 096.31) (039.69, 096.31)      /T1_2 solverssignificantlyoutperformedtheautoencoderapproach.	<|special_separator|>
-(056.69, 077.90) (294.87, 077.90) (294.87, 085.56) (056.69, 085.56)      /T1_2 SupplementaryFig.5showsthat,althoughtheICNN-basedsolver	<|special_separator|>
-(039.69, 067.15) (068.57, 067.15) (068.57, 074.81) (039.69, 074.81)      /T1_2 slightly	<|special_separator|>
-(070.31, 067.15) (126.13, 067.15) (126.13, 074.81) (070.31, 074.81)      /T1_2 outperformed	<|special_separator|>
-(127.87, 067.15) (140.67, 067.15) (140.67, 074.81) (127.87, 074.81)      /T1_2 the	<|special_separator|>
-(142.40, 067.15) (195.01, 067.15) (195.01, 074.81) (142.40, 074.81)      /T1_2 Monge-based	<|special_separator|>
-(196.75, 067.15) (246.30, 067.15) (246.30, 074.81) (196.75, 074.81)      /T1_2 counterpart,	<|special_separator|>
-(248.04, 067.15) (266.82, 067.15) (266.82, 074.81) (248.04, 074.81)      /T1_2 their	<|special_separator|>
-(268.56, 067.15) (294.89, 067.15) (294.89, 074.81) (268.56, 074.81)      /T1_2 results	<|special_separator|>
-(039.69, 056.40) (294.78, 056.40) (294.78, 064.06) (039.69, 064.06)      /T1_2 werehighlycorrelated.Moreover,theyattainedalmostidenticalper-	<|special_separator|>
-(039.69, 045.65) (076.61, 045.65) (076.61, 053.31) (039.69, 053.31)      /T1_2 formance	<|special_separator|>
-(078.25, 045.65) (089.73, 045.65) (089.73, 053.31) (078.25, 053.31)      /T1_2 for	<|special_separator|>
-(091.36, 045.65) (103.94, 045.65) (103.94, 053.31) (091.36, 053.31)      /T1_2 the	<|special_separator|>
-(105.58, 045.65) (151.24, 045.65) (151.24, 053.31) (105.58, 053.31)      /T1_2 Wasserstein	<|special_separator|>
-(152.88, 045.65) (187.46, 045.65) (187.46, 053.31) (152.88, 053.31)      /T1_2 distance.	<|special_separator|>
-(189.10, 045.65) (220.43, 045.65) (220.43, 053.31) (189.10, 053.31)      /T1_2 Because	<|special_separator|>
-(222.06, 045.65) (230.05, 045.65) (230.05, 053.31) (222.06, 053.31)      /T1_2 of	<|special_separator|>
-(231.68, 045.65) (245.99, 045.65) (245.99, 053.31) (231.68, 053.31)      /T1_2 this	<|special_separator|>
-(247.63, 045.65) (262.00, 045.65) (262.00, 053.31) (247.63, 053.31)      /T1_2 and	<|special_separator|>
-(263.64, 045.65) (294.77, 045.65) (294.77, 053.31) (263.64, 053.31)      /T1_2 because	<|special_separator|>
-(306.14, 733.65) (330.65, 733.65) (330.65, 741.31) (306.14, 741.31)      /T1_2 weare	<|special_separator|>
-(332.34, 733.65) (373.21, 733.65) (373.21, 741.31) (332.34, 741.31)      /T1_2 expanding	<|special_separator|>
-(374.90, 733.65) (395.19, 733.65) (395.19, 741.31) (374.90, 741.31)      /T1_2 upon	<|special_separator|>
-(396.88, 733.65) (409.57, 733.65) (409.57, 741.31) (396.88, 741.31)      /T1_2 the	<|special_separator|>
-(411.26, 733.65) (463.40, 733.65) (463.40, 741.31) (411.26, 741.31)      /T1_2 Monge-based	<|special_separator|>
-(465.09, 733.65) (519.59, 733.65) (519.59, 741.31) (465.09, 741.31)      /T1_2 methodology,	<|special_separator|>
-(521.29, 733.65) (546.35, 733.65) (546.35, 741.31) (521.29, 741.31)      /T1_2 wedid	<|special_separator|>
-(548.04, 733.65) (561.33, 733.65) (561.33, 741.31) (548.04, 741.31)      /T1_2 not	<|special_separator|>
-(306.14, 722.90) (493.68, 722.90) (493.68, 730.56) (306.14, 730.56)      /T1_2 includeICNN-basedbenchmarksinthemainwork.	<|special_separator|>
-(323.15, 712.15) (382.16, 712.15) (382.16, 719.81) (323.15, 719.81)      /T1_2 Weadditionally	<|special_separator|>
-(383.77, 712.15) (407.53, 712.15) (407.53, 719.81) (383.77, 719.81)      /T1_2 tested	<|special_separator|>
-(409.14, 712.15) (434.05, 712.15) (434.05, 719.81) (409.14, 719.81)      /T1_2 vanilla	<|special_separator|>
-(435.66, 712.15) (471.03, 712.15) (471.03, 719.81) (435.66, 719.81)      /T1_2 attention	<|special_separator|>
-(472.64, 712.15) (502.25, 712.15) (502.25, 719.81) (472.64, 719.81)      /T1_2 pooling	<|special_separator|>
-(503.86, 712.15) (518.17, 712.15) (518.17, 719.81) (503.86, 719.81)      /T1_2 and	<|special_separator|>
-(519.78, 712.15) (561.39, 712.15) (561.39, 719.81) (519.78, 719.81)      /T1_2 multi-head	<|special_separator|>
-(306.14, 701.40) (561.26, 701.40) (561.26, 709.06) (306.14, 709.06)      /T1_2 attentionpoolingwiththe4idataasthenumberofcombinationcon-	<|special_separator|>
-(306.14, 690.65) (333.31, 690.65) (333.31, 698.31) (306.14, 698.31)      /T1_2 ditions	<|special_separator|>
-(335.01, 690.65) (343.15, 690.65) (343.15, 698.31) (335.01, 698.31)      /T1_2 as	<|special_separator|>
-(344.85, 690.65) (367.73, 690.65) (367.73, 698.31) (344.85, 698.31)      /T1_2 small.	<|special_separator|>
-(369.42, 690.65) (378.78, 690.65) (378.78, 698.31) (369.42, 698.31)      /T1_2 As	<|special_separator|>
-(380.47, 690.65) (483.54, 690.65) (483.54, 698.31) (380.47, 698.31)      /T1_2 weshowinSupplementary	<|special_separator|>
-(485.24, 690.65) (506.33, 690.65) (506.33, 698.31) (485.24, 698.31)      /T1_2 Table	<|special_separator|>
-(508.03, 690.65) (514.76, 690.65) (514.76, 698.31) (508.03, 698.31)      /T1_2 9,	<|special_separator|>
-(516.46, 690.65) (530.94, 690.65) (530.94, 698.31) (516.46, 698.31)      /T1_2 this	<|special_separator|>
-(532.64, 690.65) (561.19, 690.65) (561.19, 698.31) (532.64, 698.31)      /T1_2 yielded	<|special_separator|>
-(306.14, 679.90) (561.22, 679.90) (561.22, 687.56) (306.14, 687.56)      /T1_2 marginalbenefitsoverthe4idatasetonlyuponcarefulregularization.	<|special_separator|>
-(306.14, 658.44) (388.66, 658.44) (388.66, 666.42) (306.14, 666.42)      /T1_5 Reportingsummary	<|special_separator|>
-(306.14, 647.65) (336.03, 647.65) (336.03, 655.31) (306.14, 655.31)      /T1_2 Further	<|special_separator|>
-(338.04, 647.65) (385.90, 647.65) (385.90, 655.31) (338.04, 655.31)      /T1_2 information	<|special_separator|>
-(387.91, 647.65) (398.14, 647.65) (398.14, 655.31) (387.91, 655.31)      /T1_2 on	<|special_separator|>
-(400.15, 647.65) (434.02, 647.65) (434.02, 655.31) (400.15, 655.31)      /T1_2 research	<|special_separator|>
-(436.03, 647.65) (462.37, 647.65) (462.37, 655.31) (436.03, 655.31)      /T1_2 design	<|special_separator|>
-(464.38, 647.65) (470.71, 647.65) (470.71, 655.31) (464.38, 655.31)      /T1_2 is	<|special_separator|>
-(472.72, 647.65) (507.76, 647.65) (507.76, 655.31) (472.72, 655.31)      /T1_2 available	<|special_separator|>
-(509.77, 647.65) (517.31, 647.65) (517.31, 655.31) (509.77, 655.31)      /T1_2 in	<|special_separator|>
-(519.32, 647.65) (532.27, 647.65) (532.27, 655.31) (519.32, 655.31)      /T1_2 the	<|special_separator|>
-(534.28, 647.65) (561.39, 647.65) (561.39, 655.31) (534.28, 655.31)      /T1_2 Nature	<|special_separator|>
-(306.14, 636.90) (494.72, 636.90) (494.72, 644.56) (306.14, 644.56)      /T1_2 PortfolioReportingSummarylinkedtothisarticle.	<|special_separator|>
-(306.14, 615.10) (389.87, 615.10) (389.87, 624.86) (306.14, 624.86)      /T1_7 Dataavailability	<|special_separator|>
-(306.14, 604.65) (324.58, 604.65) (324.58, 612.31) (306.14, 612.31)      /T1_2 Both	<|special_separator|>
-(326.22, 604.65) (358.23, 604.65) (358.23, 612.31) (326.22, 612.31)      /T1_2 datasets	<|special_separator|>
-(359.87, 604.65) (378.08, 604.65) (378.08, 612.31) (359.87, 612.31)      /T1_2 used	<|special_separator|>
-(379.72, 604.65) (387.06, 604.65) (387.06, 612.31) (379.72, 612.31)      /T1_2 in	<|special_separator|>
-(388.70, 604.65) (403.01, 604.65) (403.01, 612.31) (388.70, 612.31)      /T1_2 this	<|special_separator|>
-(404.66, 604.65) (426.05, 604.65) (426.05, 612.31) (404.66, 612.31)      /T1_2 study	<|special_separator|>
-(427.69, 604.65) (439.75, 604.65) (439.75, 612.31) (427.69, 612.31)      /T1_2 are	<|special_separator|>
-(441.39, 604.65) (472.35, 604.65) (472.35, 612.31) (441.39, 612.31)      /T1_2 publicly	<|special_separator|>
-(473.99, 604.65) (510.05, 604.65) (510.05, 612.31) (473.99, 612.31)      /T1_2 available.	<|special_separator|>
-(511.69, 604.65) (526.28, 604.65) (526.28, 612.31) (511.69, 612.31)      /T1_2 The	<|special_separator|>
-(527.92, 604.65) (561.26, 604.65) (561.26, 612.31) (527.92, 612.31)      /T1_2 preproc-	<|special_separator|>
-(306.14, 593.90) (561.29, 593.90) (561.29, 601.56) (306.14, 601.56)      /T1_2 essedversionsoftheSciPlex3and4idatasetshavebeenmadeavailable	<|special_separator|>
-(306.14, 583.15) (329.63, 583.15) (329.63, 590.81) (306.14, 590.81)      /T1_2 under	<|special_separator|>
-(331.40, 583.15) (335.76, 583.15) (335.76, 590.81) (331.40, 590.81)      /T1_2 a	<|special_separator|>
-(337.52, 583.15) (380.98, 583.15) (380.98, 590.81) (337.52, 590.81)      /T1_2 permanent	<|special_separator|>
-(382.74, 583.15) (397.96, 583.15) (397.96, 590.81) (382.74, 590.81)      /T1_2 DOI	<|special_separator|>
-(399.72, 583.15) (414.42, 583.15) (414.42, 590.81) (399.72, 590.81)      /T1_2 and	<|special_separator|>
-(416.18, 583.15) (430.01, 583.15) (430.01, 590.81) (416.18, 590.81)      /T1_2 can	<|special_separator|>
-(431.77, 583.15) (441.43, 583.15) (441.43, 590.81) (431.77, 590.81)      /T1_2 be	<|special_separator|>
-(443.19, 583.15) (491.95, 583.15) (491.95, 590.81) (443.19, 590.81)      /T1_2 downloaded	<|special_separator|>
-(493.71, 583.15) (513.03, 583.15) (513.03, 590.81) (493.71, 590.81)      /T1_2 from	<|special_separator|>
-(514.79, 583.15) (561.25, 583.15) (561.25, 590.81) (514.79, 590.81)      /T1_2 http://www.	<|special_separator|>
-(306.14, 572.40) (561.20, 572.40) (561.20, 580.06) (306.14, 580.06)      /T1_2 research-collection.ethz.ch/entities/researchdata/7964ba4a-92c7-	<|special_separator|>
-(306.14, 561.65) (433.18, 561.65) (433.18, 569.31) (306.14, 569.31)      /T1_2 4031-80f4-358f681c5056(ref.66).	<|special_separator|>
-(306.14, 539.85) (391.75, 539.85) (391.75, 549.61) (306.14, 549.61)      /T1_7 Codeavailability	<|special_separator|>
-(306.14, 529.40) (321.11, 529.40) (321.11, 537.06) (306.14, 537.06)      /T1_2 The	<|special_separator|>
-(323.08, 529.40) (349.64, 529.40) (349.64, 537.06) (323.08, 537.06)      /T1_2 source	<|special_separator|>
-(351.61, 529.40) (370.92, 529.40) (370.92, 537.06) (351.61, 537.06)      /T1_2 code	<|special_separator|>
-(372.89, 529.40) (384.71, 529.40) (384.71, 537.06) (372.89, 537.06)      /T1_2 for	<|special_separator|>
-(386.68, 529.40) (436.05, 529.40) (436.05, 537.06) (386.68, 537.06)      /T1_2 reproducing	<|special_separator|>
-(438.01, 529.40) (450.95, 529.40) (450.95, 537.06) (438.01, 537.06)      /T1_2 the	<|special_separator|>
-(452.92, 529.40) (502.60, 529.40) (502.60, 537.06) (452.92, 537.06)      /T1_2 experiments	<|special_separator|>
-(504.57, 529.40) (510.89, 529.40) (510.89, 537.06) (504.57, 537.06)      /T1_2 is	<|special_separator|>
-(512.86, 529.40) (547.83, 529.40) (547.83, 537.06) (512.86, 537.06)      /T1_2 available	<|special_separator|>
-(549.80, 529.40) (561.24, 529.40) (561.24, 537.06) (549.80, 537.06)      /T1_2 via	<|special_separator|>
-(306.14, 518.65) (336.55, 518.65) (336.55, 526.31) (306.14, 526.31)      /T1_2 Zenodo	<|special_separator|>
-(338.51, 518.65) (346.20, 518.65) (346.20, 526.31) (338.51, 526.31)      /T1_2 at	<|special_separator|>
-(348.16, 518.65) (510.11, 518.65) (510.11, 526.31) (348.16, 526.31)      /T1_2 https://doi.org/10.5281/zenodo.17854338	<|special_separator|>
-(512.07, 518.65) (528.42, 518.65) (528.42, 526.31) (512.07, 526.31)      /T1_2 (ref.	<|special_separator|>
-(530.39, 518.65) (545.13, 518.65) (545.13, 526.31) (530.39, 526.31)      /T1_2 67).	<|special_separator|>
-(547.09, 518.65) (552.98, 518.65) (552.98, 526.31) (547.09, 526.31)      /T1_2 It	<|special_separator|>
-(554.94, 518.65) (561.26, 518.65) (561.26, 526.31) (554.94, 526.31)      /T1_2 is	<|special_separator|>
-(306.14, 507.90) (561.28, 507.90) (561.28, 515.56) (306.14, 515.56)      /T1_2 alsoavailableunderanMITlicenceviaGitHubathttps://github.com/	<|special_separator|>
-(306.14, 497.15) (413.99, 497.15) (413.99, 504.81) (306.14, 504.81)      /T1_2 AI4SCR/Conditional-Monge	<|special_separator|>
-(415.70, 497.15) (430.27, 497.15) (430.27, 504.81) (415.70, 504.81)      /T1_2 and	<|special_separator|>
-(431.98, 497.15) (445.69, 497.15) (445.69, 504.81) (431.98, 504.81)      /T1_2 can	<|special_separator|>
-(447.40, 497.15) (456.98, 497.15) (456.98, 504.81) (447.40, 504.81)      /T1_2 be	<|special_separator|>
-(458.69, 497.15) (492.11, 497.15) (492.11, 504.81) (458.69, 504.81)      /T1_2 installed	<|special_separator|>
-(493.83, 497.15) (512.97, 497.15) (512.97, 504.81) (493.83, 504.81)      /T1_2 from	<|special_separator|>
-(514.68, 497.15) (531.62, 497.15) (531.62, 504.81) (514.68, 504.81)      /T1_2 PyPI	<|special_separator|>
-(533.34, 497.15) (561.31, 497.15) (561.31, 504.81) (533.34, 504.81)      /T1_2 (http://	<|special_separator|>
-(306.14, 486.40) (451.90, 486.40) (451.90, 494.06) (306.14, 494.06)      /T1_2 pypi.org/project/cmonge/)ascmonge.	<|special_separator|>
-(306.14, 464.60) (362.82, 464.60) (362.82, 474.36) (306.14, 474.36)      /T1_7 References	<|special_separator|>
-(306.14, 454.56) (311.33, 454.56) (311.33, 461.57) (306.14, 461.57)      /T1_8 1.	<|special_separator|>
-(323.15, 454.56) (357.24, 454.56) (357.24, 461.57) (323.15, 461.57)      /T1_8 Frangieh,	<|special_separator|>
-(358.84, 454.56) (366.60, 454.56) (366.60, 461.57) (358.84, 461.57)      /T1_8 C.	<|special_separator|>
-(368.20, 454.56) (373.47, 454.56) (373.47, 461.57) (368.20, 461.57)      /T1_8 J.	<|special_separator|>
-(375.07, 454.56) (382.35, 454.56) (382.35, 461.57) (375.07, 461.57)      /T1_8 et	<|special_separator|>
-(383.95, 454.56) (392.67, 454.56) (392.67, 461.57) (383.95, 461.57)      /T1_8 al.	<|special_separator|>
-(394.27, 454.56) (435.95, 454.56) (435.95, 461.57) (394.27, 461.57)      /T1_8 Multimodal	<|special_separator|>
-(437.55, 454.56) (463.61, 454.56) (463.61, 461.57) (437.55, 461.57)      /T1_8 pooled	<|special_separator|>
-(465.21, 454.56) (524.60, 454.56) (524.60, 461.57) (465.21, 461.57)      /T1_8 perturb-cite-seq	<|special_separator|>
-(526.20, 454.56) (554.58, 454.56) (554.58, 461.57) (526.20, 461.57)      /T1_8 screens	<|special_separator|>
-(323.15, 443.80) (329.87, 443.80) (329.87, 450.82) (323.15, 450.82)      /T1_8 in	<|special_separator|>
-(331.47, 443.80) (357.23, 443.80) (357.23, 450.82) (331.47, 450.82)      /T1_8 patient	<|special_separator|>
-(358.83, 443.80) (386.17, 443.80) (386.17, 450.82) (358.83, 450.82)      /T1_8 models	<|special_separator|>
-(387.77, 443.80) (410.82, 443.80) (410.82, 450.82) (387.77, 450.82)      /T1_8 define	<|special_separator|>
-(412.42, 443.80) (458.76, 443.80) (458.76, 450.82) (412.42, 450.82)      /T1_8 mechanisms	<|special_separator|>
-(460.35, 443.80) (467.85, 443.80) (467.85, 450.82) (460.35, 450.82)      /T1_8 of	<|special_separator|>
-(469.45, 443.80) (494.70, 443.80) (494.70, 450.82) (469.45, 450.82)      /T1_8 cancer	<|special_separator|>
-(496.30, 443.80) (558.15, 443.80) (558.15, 450.82) (496.30, 450.82)      /T1_8 immuneevasion.	<|special_separator|>
-(323.15, 433.05) (337.74, 433.05) (337.74, 440.07) (323.15, 440.07)      /T1_9 Nat.	<|special_separator|>
-(339.26, 433.05) (363.32, 433.05) (363.32, 440.07) (339.26, 440.07)      /T1_9 Genet.	<|special_separator|>
-(364.90, 433.07) (374.77, 433.07) (374.77, 440.07) (364.90, 440.07)      /T1_6 53	<|special_separator|>
-(374.63, 433.05) (376.89, 433.05) (376.89, 440.07) (374.63, 440.07)      /T1_8 ,	<|special_separator|>
-(378.49, 433.05) (408.70, 433.05) (408.70, 440.07) (378.49, 440.07)      /T1_8 332-341	<|special_separator|>
-(410.30, 433.05) (434.63, 433.05) (434.63, 440.07) (410.30, 440.07)      /T1_8 (2021).	<|special_separator|>
-(306.14, 422.30) (312.65, 422.30) (312.65, 429.31) (306.14, 429.31)      /T1_8 2.	<|special_separator|>
-(323.15, 422.30) (341.60, 422.30) (341.60, 429.31) (323.15, 429.31)      /T1_8 Dixit,	<|special_separator|>
-(343.20, 422.30) (350.82, 422.30) (350.82, 429.31) (343.20, 429.31)      /T1_8 A.	<|special_separator|>
-(352.42, 422.30) (359.70, 422.30) (359.70, 429.31) (352.42, 429.31)      /T1_8 et	<|special_separator|>
-(361.30, 422.30) (370.02, 422.30) (370.02, 429.31) (361.30, 429.31)      /T1_8 al.	<|special_separator|>
-(371.62, 422.30) (416.62, 422.30) (416.62, 429.31) (371.62, 429.31)      /T1_8 Perturb-seq:	<|special_separator|>
-(418.22, 422.30) (455.81, 422.30) (455.81, 429.31) (418.22, 429.31)      /T1_8 dissecting	<|special_separator|>
-(457.41, 422.30) (494.56, 422.30) (494.56, 429.31) (457.41, 429.31)      /T1_8 molecular	<|special_separator|>
-(496.16, 422.30) (522.95, 422.30) (522.95, 429.31) (496.16, 429.31)      /T1_8 circuits	<|special_separator|>
-(524.55, 422.30) (540.26, 422.30) (540.26, 429.31) (524.55, 429.31)      /T1_8 with	<|special_separator|>
-(323.15, 411.55) (353.71, 411.55) (353.71, 418.56) (323.15, 418.56)      /T1_8 scalable	<|special_separator|>
-(355.31, 411.55) (393.30, 411.55) (393.30, 418.56) (355.31, 418.56)      /T1_8 single-cell	<|special_separator|>
-(394.90, 411.55) (443.28, 411.55) (443.28, 418.56) (394.90, 418.56)      /T1_8 RNAprofiling	<|special_separator|>
-(444.88, 411.55) (452.38, 411.55) (452.38, 418.56) (444.88, 418.56)      /T1_8 of	<|special_separator|>
-(453.98, 411.55) (480.04, 411.55) (480.04, 418.56) (453.98, 418.56)      /T1_8 pooled	<|special_separator|>
-(481.64, 411.55) (509.34, 411.55) (509.34, 418.56) (481.64, 418.56)      /T1_8 genetic	<|special_separator|>
-(510.94, 411.55) (541.47, 411.55) (541.47, 418.56) (510.94, 418.56)      /T1_8 screens.	<|special_separator|>
-(543.07, 411.55) (557.53, 411.55) (557.53, 418.56) (543.07, 418.56)      /T1_9 Cell	<|special_separator|>
-(323.15, 400.81) (335.90, 400.81) (335.90, 407.81) (323.15, 407.81)      /T1_6 167	<|special_separator|>
-(335.76, 400.80) (338.01, 400.80) (338.01, 407.81) (335.76, 407.81)      /T1_8 ,	<|special_separator|>
-(339.61, 400.80) (378.13, 400.80) (378.13, 407.81) (339.61, 407.81)      /T1_8 1853-1866	<|special_separator|>
-(379.73, 400.80) (404.58, 400.80) (404.58, 407.81) (379.73, 407.81)      /T1_8 (2016).	<|special_separator|>
-(306.14, 390.04) (312.97, 390.04) (312.97, 397.06) (306.14, 397.06)      /T1_8 3.	<|special_separator|>
-(323.15, 390.04) (358.87, 390.04) (358.87, 397.06) (323.15, 397.06)      /T1_8 Srivatsan,	<|special_separator|>
-(360.47, 390.04) (367.40, 390.04) (367.40, 397.06) (360.47, 397.06)      /T1_8 S.	<|special_separator|>
-(369.00, 390.04) (376.10, 390.04) (376.10, 397.06) (369.00, 397.06)      /T1_8 R.	<|special_separator|>
-(377.70, 390.04) (384.98, 390.04) (384.98, 397.06) (377.70, 397.06)      /T1_8 et	<|special_separator|>
-(386.58, 390.04) (395.29, 390.04) (395.29, 397.06) (386.58, 397.06)      /T1_8 al.	<|special_separator|>
-(396.89, 390.04) (432.23, 390.04) (432.23, 397.06) (396.89, 397.06)      /T1_8 Massively	<|special_separator|>
-(433.83, 390.04) (468.26, 390.04) (468.26, 397.06) (433.83, 397.06)      /T1_8 multiplex	<|special_separator|>
-(469.86, 390.04) (503.56, 390.04) (503.56, 397.06) (469.86, 397.06)      /T1_8 chemical	<|special_separator|>
-(323.15, 379.29) (380.47, 379.29) (380.47, 386.31) (323.15, 386.31)      /T1_8 transcriptomics	<|special_separator|>
-(382.07, 379.29) (389.12, 379.29) (389.12, 386.31) (382.07, 386.31)      /T1_8 at	<|special_separator|>
-(390.72, 379.29) (428.71, 379.29) (428.71, 386.31) (390.72, 386.31)      /T1_8 single-cell	<|special_separator|>
-(430.31, 379.29) (469.27, 379.29) (469.27, 386.31) (430.31, 386.31)      /T1_8 resolution.	<|special_separator|>
-(470.87, 379.29) (499.51, 379.29) (499.51, 386.31) (470.87, 386.31)      /T1_9 Science	<|special_separator|>
-(501.17, 379.31) (515.56, 379.31) (515.56, 386.31) (501.17, 386.31)      /T1_6 367	<|special_separator|>
-(515.41, 379.29) (517.67, 379.29) (517.67, 386.31) (515.41, 386.31)      /T1_8 ,	<|special_separator|>
-(519.27, 379.29) (540.58, 379.29) (540.58, 386.31) (519.27, 386.31)      /T1_8 45-51	<|special_separator|>
-(323.15, 368.54) (349.70, 368.54) (349.70, 375.55) (323.15, 375.55)      /T1_8 (2020).	<|special_separator|>
-(306.14, 357.79) (313.15, 357.79) (313.15, 364.80) (306.14, 364.80)      /T1_8 4.	<|special_separator|>
-(323.15, 357.79) (330.60, 357.79) (330.60, 364.80) (323.15, 364.80)      /T1_8 Ji,	<|special_separator|>
-(332.20, 357.79) (340.33, 357.79) (340.33, 364.80) (332.20, 364.80)      /T1_8 Y.,	<|special_separator|>
-(341.93, 357.79) (378.22, 357.79) (378.22, 364.80) (341.93, 364.80)      /T1_8 Lotfollahi,	<|special_separator|>
-(379.82, 357.79) (390.97, 357.79) (390.97, 364.80) (379.82, 364.80)      /T1_8 M.,	<|special_separator|>
-(392.57, 357.79) (410.98, 357.79) (410.98, 364.80) (392.57, 364.80)      /T1_8 Wolf,	<|special_separator|>
-(412.58, 357.79) (418.39, 357.79) (418.39, 364.80) (412.58, 364.80)      /T1_8 F.	<|special_separator|>
-(419.99, 357.79) (427.61, 357.79) (427.61, 364.80) (419.99, 364.80)      /T1_8 A.	<|special_separator|>
-(429.21, 357.79) (457.73, 357.79) (457.73, 364.80) (429.21, 364.80)      /T1_8 &Theis,	<|special_separator|>
-(459.32, 357.79) (465.13, 357.79) (465.13, 364.80) (459.32, 364.80)      /T1_8 F.	<|special_separator|>
-(466.73, 357.79) (472.00, 357.79) (472.00, 364.80) (466.73, 364.80)      /T1_8 J.	<|special_separator|>
-(473.60, 357.79) (504.86, 357.79) (504.86, 364.80) (473.60, 364.80)      /T1_8 Machine	<|special_separator|>
-(506.46, 357.79) (536.61, 357.79) (536.61, 364.80) (506.46, 364.80)      /T1_8 learning	<|special_separator|>
-(538.21, 357.79) (548.41, 357.79) (548.41, 364.80) (538.21, 364.80)      /T1_8 for	<|special_separator|>
-(323.15, 347.04) (375.50, 347.04) (375.50, 354.05) (323.15, 354.05)      /T1_8 perturbational	<|special_separator|>
-(377.10, 347.04) (415.10, 347.04) (415.10, 354.05) (377.10, 354.05)      /T1_8 single-cell	<|special_separator|>
-(416.70, 347.04) (440.98, 347.04) (440.98, 354.05) (416.70, 354.05)      /T1_8 omics.	<|special_separator|>
-(442.58, 347.04) (457.04, 347.04) (457.04, 354.05) (442.58, 354.05)      /T1_9 Cell	<|special_separator|>
-(458.56, 347.04) (475.80, 347.04) (475.80, 354.05) (458.56, 354.05)      /T1_9 Syst.	<|special_separator|>
-(477.45, 347.05) (485.50, 347.05) (485.50, 354.05) (477.45, 354.05)      /T1_6 12	<|special_separator|>
-(485.37, 347.04) (487.62, 347.04) (487.62, 354.05) (485.37, 354.05)      /T1_8 ,	<|special_separator|>
-(489.22, 347.04) (519.82, 347.04) (519.82, 354.05) (489.22, 354.05)      /T1_8 522-537	<|special_separator|>
-(521.42, 347.04) (545.75, 347.04) (545.75, 354.05) (521.42, 354.05)      /T1_8 (2021).	<|special_separator|>
-(306.14, 336.28) (312.67, 336.28) (312.67, 343.30) (306.14, 343.30)      /T1_8 5.	<|special_separator|>
-(323.15, 336.28) (345.30, 336.28) (345.30, 343.30) (323.15, 343.30)      /T1_8 Peidli,	<|special_separator|>
-(346.90, 336.28) (353.83, 336.28) (353.83, 343.30) (346.90, 343.30)      /T1_8 S.	<|special_separator|>
-(355.43, 336.28) (362.71, 336.28) (362.71, 343.30) (355.43, 343.30)      /T1_8 et	<|special_separator|>
-(364.31, 336.28) (373.02, 336.28) (373.02, 343.30) (364.31, 343.30)      /T1_8 al.	<|special_separator|>
-(374.62, 336.28) (412.16, 336.28) (412.16, 343.30) (374.62, 343.30)      /T1_8 scPerturb:	<|special_separator|>
-(413.76, 336.28) (457.26, 336.28) (457.26, 343.30) (413.76, 343.30)      /T1_8 harmonized	<|special_separator|>
-(458.86, 336.28) (496.85, 336.28) (496.85, 343.30) (458.86, 343.30)      /T1_8 single-cell	<|special_separator|>
-(498.45, 336.28) (544.35, 336.28) (544.35, 343.30) (498.45, 343.30)      /T1_8 perturbation	<|special_separator|>
-(323.15, 325.53) (341.53, 325.53) (341.53, 332.55) (323.15, 332.55)      /T1_8 data.	<|special_separator|>
-(343.13, 325.53) (357.72, 325.53) (357.72, 332.55) (343.13, 332.55)      /T1_9 Nat.	<|special_separator|>
-(359.24, 325.53) (390.66, 325.53) (390.66, 332.55) (359.24, 332.55)      /T1_9 Methods	<|special_separator|>
-(392.24, 325.55) (400.29, 325.55) (400.29, 332.55) (392.24, 332.55)      /T1_6 21	<|special_separator|>
-(400.16, 325.53) (402.41, 325.53) (402.41, 332.55) (400.16, 332.55)      /T1_8 ,	<|special_separator|>
-(404.01, 325.53) (435.21, 325.53) (435.21, 332.55) (404.01, 332.55)      /T1_8 531-540	<|special_separator|>
-(436.81, 325.53) (462.61, 325.53) (462.61, 332.55) (436.81, 332.55)      /T1_8 (2024).	<|special_separator|>
-(306.14, 314.78) (313.05, 314.78) (313.05, 321.79) (306.14, 321.79)      /T1_8 6.	<|special_separator|>
-(323.15, 314.78) (354.68, 314.78) (354.68, 321.79) (323.15, 321.79)      /T1_8 Ianevski,	<|special_separator|>
-(356.28, 314.78) (363.91, 314.78) (363.91, 321.79) (356.28, 321.79)      /T1_8 A.	<|special_separator|>
-(365.51, 314.78) (372.79, 314.78) (372.79, 321.79) (365.51, 321.79)      /T1_8 et	<|special_separator|>
-(374.39, 314.78) (383.10, 314.78) (383.10, 321.79) (374.39, 321.79)      /T1_8 al.	<|special_separator|>
-(384.70, 314.78) (423.87, 314.78) (423.87, 321.79) (384.70, 321.79)      /T1_8 Single-cell	<|special_separator|>
-(425.47, 314.78) (480.87, 314.78) (480.87, 321.79) (425.47, 321.79)      /T1_8 transcriptomes	<|special_separator|>
-(482.47, 314.78) (510.09, 314.78) (510.09, 321.79) (482.47, 321.79)      /T1_8 identify	<|special_separator|>
-(323.15, 304.03) (379.51, 304.03) (379.51, 311.04) (323.15, 311.04)      /T1_8 patient-tailored	<|special_separator|>
-(381.11, 304.03) (415.23, 304.03) (415.23, 311.04) (381.11, 311.04)      /T1_8 therapies	<|special_separator|>
-(416.83, 304.03) (427.03, 304.03) (427.03, 311.04) (416.83, 311.04)      /T1_8 for	<|special_separator|>
-(428.63, 304.03) (461.21, 304.03) (461.21, 311.04) (428.63, 311.04)      /T1_8 selective	<|special_separator|>
-(462.81, 304.03) (508.82, 304.03) (508.82, 311.04) (462.81, 311.04)      /T1_8 co-inhibition	<|special_separator|>
-(510.42, 304.03) (517.92, 304.03) (517.92, 311.04) (510.42, 311.04)      /T1_8 of	<|special_separator|>
-(519.52, 304.03) (544.77, 304.03) (544.77, 311.04) (519.52, 311.04)      /T1_8 cancer	<|special_separator|>
-(323.15, 293.28) (349.73, 293.28) (349.73, 300.29) (323.15, 300.29)      /T1_8 clones.	<|special_separator|>
-(351.34, 293.28) (403.29, 293.28) (403.29, 300.29) (351.34, 300.29)      /T1_9 Nat.Commun.	<|special_separator|>
-(404.87, 293.29) (413.10, 293.29) (413.10, 300.29) (404.87, 300.29)      /T1_6 15	<|special_separator|>
-(412.97, 293.28) (415.22, 293.28) (415.22, 300.29) (412.97, 300.29)      /T1_8 ,	<|special_separator|>
-(416.82, 293.28) (434.91, 293.28) (434.91, 300.29) (416.82, 300.29)      /T1_8 8579	<|special_separator|>
-(436.51, 293.28) (462.31, 293.28) (462.31, 300.29) (436.51, 300.29)      /T1_8 (2024).	<|special_separator|>
-(306.14, 282.52) (311.70, 282.52) (311.70, 289.54) (306.14, 289.54)      /T1_8 7.	<|special_separator|>
-(323.15, 282.52) (336.43, 282.52) (336.43, 289.54) (323.15, 289.54)      /T1_8 Bai,	<|special_separator|>
-(338.03, 282.52) (344.94, 282.52) (344.94, 289.54) (338.03, 289.54)      /T1_8 Z.	<|special_separator|>
-(346.54, 282.52) (353.82, 282.52) (353.82, 289.54) (346.54, 289.54)      /T1_8 et	<|special_separator|>
-(355.42, 282.52) (364.13, 282.52) (364.13, 289.54) (355.42, 289.54)      /T1_8 al.	<|special_separator|>
-(365.73, 282.52) (404.90, 282.52) (404.90, 289.54) (365.73, 289.54)      /T1_8 Single-cell	<|special_separator|>
-(406.50, 282.52) (447.37, 282.52) (447.37, 289.54) (406.50, 289.54)      /T1_8 CARTatlas	<|special_separator|>
-(448.97, 282.52) (474.99, 282.52) (474.99, 289.54) (448.97, 289.54)      /T1_8 reveals	<|special_separator|>
-(476.59, 282.52) (492.91, 282.52) (492.91, 289.54) (476.59, 289.54)      /T1_8 type	<|special_separator|>
-(494.51, 282.52) (498.94, 282.52) (498.94, 289.54) (494.51, 289.54)      /T1_8 2	<|special_separator|>
-(500.54, 282.52) (531.17, 282.52) (531.17, 289.54) (500.54, 289.54)      /T1_8 function	<|special_separator|>
-(532.77, 282.52) (539.49, 282.52) (539.49, 289.54) (532.77, 289.54)      /T1_8 in	<|special_separator|>
-(323.15, 271.77) (346.03, 271.77) (346.03, 278.79) (323.15, 278.79)      /T1_8 8-year	<|special_separator|>
-(347.63, 271.77) (385.03, 271.77) (385.03, 278.79) (347.63, 278.79)      /T1_8 leukaemia	<|special_separator|>
-(386.63, 271.77) (424.04, 271.77) (424.04, 278.79) (386.63, 278.79)      /T1_8 remission.	<|special_separator|>
-(425.64, 271.77) (449.77, 271.77) (449.77, 278.79) (425.64, 278.79)      /T1_9 Nature	<|special_separator|>
-(451.38, 271.79) (466.63, 271.79) (466.63, 278.79) (451.38, 278.79)      /T1_6 634	<|special_separator|>
-(466.49, 271.77) (468.74, 271.77) (468.74, 278.79) (466.49, 278.79)      /T1_8 ,	<|special_separator|>
-(470.34, 271.77) (497.84, 271.77) (497.84, 278.79) (470.34, 278.79)      /T1_8 702-711	<|special_separator|>
-(499.44, 271.77) (525.24, 271.77) (525.24, 278.79) (499.44, 278.79)      /T1_8 (2024).	<|special_separator|>
-(306.14, 261.02) (312.95, 261.02) (312.95, 268.03) (306.14, 268.03)      /T1_8 8.	<|special_separator|>
-(323.15, 261.02) (359.44, 261.02) (359.44, 268.03) (323.15, 268.03)      /T1_8 Lotfollahi,	<|special_separator|>
-(361.04, 261.02) (372.19, 261.02) (372.19, 268.03) (361.04, 268.03)      /T1_8 M.,	<|special_separator|>
-(373.79, 261.02) (392.20, 261.02) (392.20, 268.03) (373.79, 268.03)      /T1_8 Wolf,	<|special_separator|>
-(393.80, 261.02) (399.61, 261.02) (399.61, 268.03) (393.80, 268.03)      /T1_8 F.	<|special_separator|>
-(401.21, 261.02) (408.83, 261.02) (408.83, 268.03) (401.21, 268.03)      /T1_8 A.	<|special_separator|>
-(410.43, 261.02) (438.95, 261.02) (438.95, 268.03) (410.43, 268.03)      /T1_8 &Theis,	<|special_separator|>
-(440.55, 261.02) (446.35, 261.02) (446.35, 268.03) (440.55, 268.03)      /T1_8 F.	<|special_separator|>
-(447.95, 261.02) (453.23, 261.02) (453.23, 268.03) (447.95, 268.03)      /T1_8 J.	<|special_separator|>
-(454.83, 261.02) (478.27, 261.02) (478.27, 268.03) (454.83, 268.03)      /T1_8 scGen	<|special_separator|>
-(479.87, 261.02) (509.89, 261.02) (509.89, 268.03) (479.87, 268.03)      /T1_8 predicts	<|special_separator|>
-(511.49, 261.02) (549.48, 261.02) (549.48, 268.03) (511.49, 268.03)      /T1_8 single-cell	<|special_separator|>
-(323.15, 250.27) (369.05, 250.27) (369.05, 257.28) (323.15, 257.28)      /T1_8 perturbation	<|special_separator|>
-(370.65, 250.27) (410.18, 250.27) (410.18, 257.28) (370.65, 257.28)      /T1_8 responses.	<|special_separator|>
-(411.78, 250.27) (426.37, 250.27) (426.37, 257.28) (411.78, 257.28)      /T1_9 Nat.	<|special_separator|>
-(427.89, 250.27) (459.31, 250.27) (459.31, 257.28) (427.89, 257.28)      /T1_9 Methods	<|special_separator|>
-(460.93, 250.28) (469.48, 250.28) (469.48, 257.28) (460.93, 257.28)      /T1_6 16	<|special_separator|>
-(469.34, 250.27) (471.60, 250.27) (471.60, 257.28) (469.34, 257.28)      /T1_8 ,	<|special_separator|>
-(473.20, 250.27) (500.14, 250.27) (500.14, 257.28) (473.20, 257.28)      /T1_8 715-721	<|special_separator|>
-(501.74, 250.27) (526.52, 250.27) (526.52, 257.28) (501.74, 257.28)      /T1_8 (2019).	<|special_separator|>
-(306.14, 239.52) (312.82, 239.52) (312.82, 246.53) (306.14, 246.53)      /T1_8 9.	<|special_separator|>
-(323.15, 239.52) (359.44, 239.52) (359.44, 246.53) (323.15, 246.53)      /T1_8 Lotfollahi,	<|special_separator|>
-(361.04, 239.52) (372.19, 239.52) (372.19, 246.53) (361.04, 246.53)      /T1_8 M.,	<|special_separator|>
-(373.79, 239.52) (423.98, 239.52) (423.98, 246.53) (373.79, 246.53)      /T1_8 Naghipourfar,	<|special_separator|>
-(425.58, 239.52) (436.73, 239.52) (436.73, 246.53) (425.58, 246.53)      /T1_8 M.,	<|special_separator|>
-(438.33, 239.52) (459.77, 239.52) (459.77, 246.53) (438.33, 246.53)      /T1_8 Theis,	<|special_separator|>
-(461.37, 239.52) (467.17, 239.52) (467.17, 246.53) (461.37, 246.53)      /T1_8 F.	<|special_separator|>
-(468.77, 239.52) (474.05, 239.52) (474.05, 246.53) (468.77, 246.53)      /T1_8 J.	<|special_separator|>
-(475.65, 239.52) (501.15, 239.52) (501.15, 246.53) (475.65, 246.53)      /T1_8 &Wolf,	<|special_separator|>
-(502.75, 239.52) (508.55, 239.52) (508.55, 246.53) (502.75, 246.53)      /T1_8 F.	<|special_separator|>
-(510.15, 239.52) (517.77, 239.52) (517.77, 246.53) (510.15, 246.53)      /T1_8 A.	<|special_separator|>
-(323.15, 228.76) (365.78, 228.76) (365.78, 235.78) (323.15, 235.78)      /T1_8 Conditional	<|special_separator|>
-(367.38, 228.76) (433.81, 228.76) (433.81, 235.78) (367.38, 235.78)      /T1_8 out-of-distribution	<|special_separator|>
-(435.41, 228.76) (475.15, 228.76) (475.15, 235.78) (435.41, 235.78)      /T1_8 generation	<|special_separator|>
-(476.75, 228.76) (486.95, 228.76) (486.95, 235.78) (476.75, 235.78)      /T1_8 for	<|special_separator|>
-(488.55, 228.76) (521.19, 228.76) (521.19, 235.78) (488.55, 235.78)      /T1_8 unpaired	<|special_separator|>
-(522.79, 228.76) (539.01, 228.76) (539.01, 235.78) (522.79, 235.78)      /T1_8 data	<|special_separator|>
-(540.61, 228.76) (560.63, 228.76) (560.63, 235.78) (540.61, 235.78)      /T1_8 using	<|special_separator|>
-(323.15, 218.01) (351.30, 218.01) (351.30, 225.03) (323.15, 225.03)      /T1_8 transfer	<|special_separator|>
-(352.90, 218.01) (369.64, 218.01) (369.64, 225.03) (352.90, 225.03)      /T1_8 VAE.	<|special_separator|>
-(371.24, 218.01) (422.73, 218.01) (422.73, 225.03) (371.24, 225.03)      /T1_9 Bioinformatics	<|special_separator|>
-(424.33, 218.03) (434.51, 218.03) (434.51, 225.03) (424.33, 225.03)      /T1_6 36	<|special_separator|>
-(434.37, 218.01) (436.63, 218.01) (436.63, 225.03) (434.37, 225.03)      /T1_8 ,	<|special_separator|>
-(438.23, 218.01) (471.62, 218.01) (471.62, 225.03) (438.23, 225.03)      /T1_8 i610-i617	<|special_separator|>
-(473.22, 218.01) (499.77, 218.01) (499.77, 225.03) (473.22, 225.03)      /T1_8 (2020).	<|special_separator|>
-(306.14, 207.26) (316.43, 207.26) (316.43, 214.27) (306.14, 214.27)      /T1_8 10.	<|special_separator|>
-(323.15, 207.26) (359.44, 207.26) (359.44, 214.27) (323.15, 214.27)      /T1_8 Lotfollahi,	<|special_separator|>
-(361.04, 207.26) (370.04, 207.26) (370.04, 214.27) (361.04, 214.27)      /T1_8 M.	<|special_separator|>
-(371.64, 207.26) (378.92, 207.26) (378.92, 214.27) (371.64, 214.27)      /T1_8 et	<|special_separator|>
-(380.52, 207.26) (389.23, 207.26) (389.23, 214.27) (380.52, 214.27)      /T1_8 al.	<|special_separator|>
-(390.83, 207.26) (428.27, 207.26) (428.27, 214.27) (390.83, 214.27)      /T1_8 Predicting	<|special_separator|>
-(429.87, 207.26) (457.36, 207.26) (457.36, 214.27) (429.87, 214.27)      /T1_8 cellular	<|special_separator|>
-(458.96, 207.26) (496.35, 207.26) (496.35, 214.27) (458.96, 214.27)      /T1_8 responses	<|special_separator|>
-(497.95, 207.26) (505.51, 207.26) (505.51, 214.27) (497.95, 214.27)      /T1_8 to	<|special_separator|>
-(507.11, 207.26) (539.17, 207.26) (539.17, 214.27) (507.11, 214.27)      /T1_8 complex	<|special_separator|>
-(323.15, 196.51) (372.78, 196.51) (372.78, 203.52) (323.15, 203.52)      /T1_8 perturbations	<|special_separator|>
-(374.38, 196.51) (381.10, 196.51) (381.10, 203.52) (374.38, 203.52)      /T1_8 in	<|special_separator|>
-(382.70, 196.51) (443.05, 196.51) (443.05, 203.52) (382.70, 203.52)      /T1_8 high-throughput	<|special_separator|>
-(444.65, 196.51) (475.18, 196.51) (475.18, 203.52) (444.65, 203.52)      /T1_8 screens.	<|special_separator|>
-(476.78, 196.51) (492.54, 196.51) (492.54, 203.52) (476.78, 203.52)      /T1_9 Mol.	<|special_separator|>
-(494.06, 196.51) (511.31, 196.51) (511.31, 203.52) (494.06, 203.52)      /T1_9 Syst.	<|special_separator|>
-(512.83, 196.51) (528.62, 196.51) (528.62, 203.52) (512.83, 203.52)      /T1_9 Biol.	<|special_separator|>
-(530.28, 196.52) (538.79, 196.52) (538.79, 203.52) (530.28, 203.52)      /T1_6 19	<|special_separator|>
-(538.65, 196.51) (540.91, 196.51) (540.91, 203.52) (538.65, 203.52)      /T1_8 ,	<|special_separator|>
-(323.15, 185.76) (376.03, 185.76) (376.03, 192.77) (323.15, 192.77)      /T1_8 MSB202211517	<|special_separator|>
-(377.63, 185.76) (403.60, 185.76) (403.60, 192.77) (377.63, 192.77)      /T1_8 (2023).	<|special_separator|>
-(306.14, 175.00) (314.42, 175.00) (314.42, 182.02) (306.14, 182.02)      /T1_8 11.	<|special_separator|>
-(323.15, 175.00) (348.76, 175.00) (348.76, 182.02) (323.15, 182.02)      /T1_8 Hetzel,	<|special_separator|>
-(350.36, 175.00) (357.00, 175.00) (357.00, 182.02) (350.36, 182.02)      /T1_8 L.	<|special_separator|>
-(358.60, 175.00) (365.88, 175.00) (365.88, 182.02) (358.60, 182.02)      /T1_8 et	<|special_separator|>
-(367.48, 175.00) (376.19, 175.00) (376.19, 182.02) (367.48, 182.02)      /T1_8 al.	<|special_separator|>
-(377.79, 175.00) (415.23, 175.00) (415.23, 182.02) (377.79, 182.02)      /T1_8 Predicting	<|special_separator|>
-(416.83, 175.00) (444.32, 175.00) (444.32, 182.02) (416.83, 182.02)      /T1_8 cellular	<|special_separator|>
-(445.92, 175.00) (483.31, 175.00) (483.31, 182.02) (445.92, 182.02)      /T1_8 responses	<|special_separator|>
-(484.91, 175.00) (492.47, 175.00) (492.47, 182.02) (484.91, 182.02)      /T1_8 to	<|special_separator|>
-(494.07, 175.00) (514.20, 175.00) (514.20, 182.02) (494.07, 182.02)      /T1_8 novel	<|special_separator|>
-(515.80, 175.00) (533.28, 175.00) (533.28, 182.02) (515.80, 182.02)      /T1_8 drug	<|special_separator|>
-(323.15, 164.25) (372.78, 164.25) (372.78, 171.27) (323.15, 171.27)      /T1_8 perturbations	<|special_separator|>
-(374.38, 164.25) (381.44, 164.25) (381.44, 171.27) (374.38, 171.27)      /T1_8 at	<|special_separator|>
-(383.04, 164.25) (387.36, 164.25) (387.36, 171.27) (383.04, 171.27)      /T1_8 a	<|special_separator|>
-(388.96, 164.25) (426.95, 164.25) (426.95, 171.27) (388.96, 171.27)      /T1_8 single-cell	<|special_separator|>
-(428.55, 164.25) (467.51, 164.25) (467.51, 171.27) (428.55, 171.27)      /T1_8 resolution.	<|special_separator|>
-(469.11, 164.25) (476.01, 164.25) (476.01, 171.27) (469.11, 171.27)      /T1_8 In	<|special_separator|>
-(477.61, 164.25) (494.44, 164.25) (494.44, 171.27) (477.61, 171.27)      /T1_9 36th	<|special_separator|>
-(495.96, 164.25) (548.97, 164.25) (548.97, 171.27) (495.96, 171.27)      /T1_9 Conferenceon	<|special_separator|>
-(323.15, 153.50) (346.90, 153.50) (346.90, 160.51) (323.15, 160.51)      /T1_9 Neural	<|special_separator|>
-(348.42, 153.50) (389.94, 153.50) (389.94, 160.51) (348.42, 160.51)      /T1_9 Information	<|special_separator|>
-(391.46, 153.50) (430.74, 153.50) (430.74, 160.51) (391.46, 160.51)      /T1_9 Processing	<|special_separator|>
-(432.26, 153.50) (462.37, 153.50) (462.37, 160.51) (432.26, 160.51)      /T1_9 Systems	<|special_separator|>
-(463.89, 153.50) (495.22, 153.50) (495.22, 160.51) (463.89, 160.51)      /T1_9 (NeurIPS	<|special_separator|>
-(496.74, 153.50) (517.57, 153.50) (517.57, 160.51) (496.74, 160.51)      /T1_9 2022)	<|special_separator|>
-(519.21, 153.50) (545.87, 153.50) (545.87, 160.51) (519.21, 160.51)      /T1_8 https://	<|special_separator|>
-(323.15, 142.75) (454.82, 142.75) (454.82, 149.76) (323.15, 149.76)      /T1_8 openreview.net/pdf?id=vRrFVHxFiXJ	<|special_separator|>
-(456.42, 142.75) (481.99, 142.75) (481.99, 149.76) (456.42, 149.76)      /T1_8 (2022).	<|special_separator|>
-(306.14, 132.00) (315.74, 132.00) (315.74, 139.01) (306.14, 139.01)      /T1_8 12.	<|special_separator|>
-(323.15, 132.00) (334.38, 132.00) (334.38, 139.01) (323.15, 139.01)      /T1_8 Yu,	<|special_separator|>
-(335.98, 132.00) (346.12, 132.00) (346.12, 139.01) (335.98, 139.01)      /T1_8 H.,	<|special_separator|>
-(347.72, 132.00) (367.19, 132.00) (367.19, 139.01) (347.72, 139.01)      /T1_8 Qian,	<|special_separator|>
-(368.79, 132.00) (379.88, 132.00) (379.88, 139.01) (368.79, 139.01)      /T1_8 W.,	<|special_separator|>
-(381.48, 132.00) (402.97, 132.00) (402.97, 139.01) (381.48, 139.01)      /T1_8 Song,	<|special_separator|>
-(404.57, 132.00) (410.55, 132.00) (410.55, 139.01) (404.57, 139.01)      /T1_8 Y.	<|special_separator|>
-(412.15, 132.00) (444.31, 132.00) (444.31, 139.01) (412.15, 139.01)      /T1_8 &Welch,	<|special_separator|>
-(445.91, 132.00) (451.19, 132.00) (451.19, 139.01) (445.91, 139.01)      /T1_8 J.	<|special_separator|>
-(452.79, 132.00) (460.03, 132.00) (460.03, 139.01) (452.79, 139.01)      /T1_8 D.	<|special_separator|>
-(461.63, 132.00) (501.61, 132.00) (501.61, 139.01) (461.63, 139.01)      /T1_8 PerturbNet	<|special_separator|>
-(503.21, 132.00) (533.23, 132.00) (533.23, 139.01) (503.21, 139.01)      /T1_8 predicts	<|special_separator|>
-(323.15, 121.24) (361.14, 121.24) (361.14, 128.26) (323.15, 128.26)      /T1_8 single-cell	<|special_separator|>
-(362.74, 121.24) (400.13, 121.24) (400.13, 128.26) (362.74, 128.26)      /T1_8 responses	<|special_separator|>
-(401.73, 121.24) (409.29, 121.24) (409.29, 128.26) (401.73, 128.26)      /T1_8 to	<|special_separator|>
-(410.89, 121.24) (437.62, 121.24) (437.62, 128.26) (410.89, 128.26)      /T1_8 unseen	<|special_separator|>
-(439.22, 121.24) (472.93, 121.24) (472.93, 128.26) (439.22, 128.26)      /T1_8 chemical	<|special_separator|>
-(474.53, 121.24) (488.44, 121.24) (488.44, 128.26) (474.53, 128.26)      /T1_8 and	<|special_separator|>
-(490.04, 121.24) (517.74, 121.24) (517.74, 128.26) (490.04, 128.26)      /T1_8 genetic	<|special_separator|>
-(323.15, 110.49) (374.94, 110.49) (374.94, 117.51) (323.15, 117.51)      /T1_8 perturbations.	<|special_separator|>
-(376.57, 110.49) (392.33, 110.49) (392.33, 117.51) (376.57, 117.51)      /T1_9 Mol.	<|special_separator|>
-(393.85, 110.49) (411.09, 110.49) (411.09, 117.51) (393.85, 117.51)      /T1_9 Syst.	<|special_separator|>
-(412.61, 110.49) (428.40, 110.49) (428.40, 117.51) (412.61, 117.51)      /T1_9 Biol.	<|special_separator|>
-(429.98, 110.51) (438.04, 110.51) (438.04, 117.51) (429.98, 117.51)      /T1_6 21	<|special_separator|>
-(437.90, 110.49) (440.16, 110.49) (440.16, 117.51) (437.90, 117.51)      /T1_8 ,	<|special_separator|>
-(441.76, 110.49) (474.72, 110.49) (474.72, 117.51) (441.76, 117.51)      /T1_8 960-982	<|special_separator|>
-(476.32, 110.49) (502.16, 110.49) (502.16, 117.51) (476.32, 117.51)      /T1_8 (2025).	<|special_separator|>
-(306.14, 099.74) (316.06, 099.74) (316.06, 106.75) (306.14, 106.75)      /T1_8 13.	<|special_separator|>
-(323.15, 099.74) (355.18, 099.74) (355.18, 106.75) (323.15, 106.75)      /T1_8 Roohani,	<|special_separator|>
-(356.78, 099.74) (364.91, 099.74) (364.91, 106.75) (356.78, 106.75)      /T1_8 Y.,	<|special_separator|>
-(366.51, 099.74) (392.96, 099.74) (392.96, 106.75) (366.51, 106.75)      /T1_8 Huang,	<|special_separator|>
-(394.56, 099.74) (401.64, 099.74) (401.64, 106.75) (394.56, 106.75)      /T1_8 K.	<|special_separator|>
-(403.24, 099.74) (446.24, 099.74) (446.24, 106.75) (403.24, 106.75)      /T1_8 &Leskovec,	<|special_separator|>
-(447.84, 099.74) (453.12, 099.74) (453.12, 106.75) (447.84, 106.75)      /T1_8 J.	<|special_separator|>
-(454.72, 099.74) (492.16, 099.74) (492.16, 106.75) (454.72, 106.75)      /T1_8 Predicting	<|special_separator|>
-(493.76, 099.74) (546.85, 099.74) (546.85, 106.75) (493.76, 106.75)      /T1_8 transcriptional	<|special_separator|>
-(323.15, 088.99) (359.95, 088.99) (359.95, 096.00) (323.15, 096.00)      /T1_8 outcomes	<|special_separator|>
-(361.55, 088.99) (369.04, 088.99) (369.04, 096.00) (361.55, 096.00)      /T1_8 of	<|special_separator|>
-(370.64, 088.99) (390.77, 088.99) (390.77, 096.00) (370.64, 096.00)      /T1_8 novel	<|special_separator|>
-(392.37, 088.99) (429.68, 088.99) (429.68, 096.00) (392.37, 096.00)      /T1_8 multigene	<|special_separator|>
-(431.28, 088.99) (480.91, 088.99) (480.91, 096.00) (431.28, 096.00)      /T1_8 perturbations	<|special_separator|>
-(482.51, 088.99) (498.22, 088.99) (498.22, 096.00) (482.51, 096.00)      /T1_8 with	<|special_separator|>
-(499.82, 088.99) (527.33, 088.99) (527.33, 096.00) (499.82, 096.00)      /T1_8 GEARS.	<|special_separator|>
-(528.93, 088.99) (543.52, 088.99) (543.52, 096.00) (528.93, 096.00)      /T1_9 Nat.	<|special_separator|>
-(323.15, 078.24) (363.82, 078.24) (363.82, 085.25) (323.15, 085.25)      /T1_9 Biotechnol.	<|special_separator|>
-(365.40, 078.25) (375.25, 078.25) (375.25, 085.25) (365.40, 085.25)      /T1_6 42	<|special_separator|>
-(375.12, 078.24) (377.37, 078.24) (377.37, 085.25) (375.12, 085.25)      /T1_8 ,	<|special_separator|>
-(378.97, 078.24) (410.09, 078.24) (410.09, 085.25) (378.97, 085.25)      /T1_8 927-935	<|special_separator|>
-(411.69, 078.24) (437.49, 078.24) (437.49, 085.25) (411.69, 085.25)      /T1_8 (2024).	<|special_separator|>
-(306.14, 067.48) (316.08, 067.48) (316.08, 074.50) (306.14, 074.50)      /T1_8 14.	<|special_separator|>
-(323.15, 067.48) (348.77, 067.48) (348.77, 074.50) (323.15, 074.50)      /T1_8 Zheng,	<|special_separator|>
-(350.37, 067.48) (356.35, 067.48) (356.35, 074.50) (350.37, 074.50)      /T1_8 Y.	<|special_separator|>
-(357.95, 067.48) (365.23, 067.48) (365.23, 074.50) (357.95, 074.50)      /T1_8 et	<|special_separator|>
-(366.83, 067.48) (375.54, 067.48) (375.54, 074.50) (366.83, 074.50)      /T1_8 al.	<|special_separator|>
-(377.14, 067.48) (443.37, 067.48) (443.37, 074.50) (377.14, 074.50)      /T1_8 Adeepgenerative	<|special_separator|>
-(444.97, 067.48) (468.57, 067.48) (468.57, 074.50) (444.97, 074.50)      /T1_8 model	<|special_separator|>
-(470.17, 067.48) (480.37, 067.48) (480.37, 074.50) (470.17, 074.50)      /T1_8 for	<|special_separator|>
-(481.97, 067.48) (526.28, 067.48) (526.28, 074.50) (481.97, 074.50)      /T1_8 deciphering	<|special_separator|>
-(527.88, 067.48) (555.37, 067.48) (555.37, 074.50) (527.88, 074.50)      /T1_8 cellular	<|special_separator|>
-(323.15, 056.73) (358.49, 056.73) (358.49, 063.75) (323.15, 063.75)      /T1_8 dynamics	<|special_separator|>
-(360.09, 056.73) (374.01, 056.73) (374.01, 063.75) (360.09, 063.75)      /T1_8 and	<|special_separator|>
-(375.61, 056.73) (382.33, 056.73) (382.33, 063.75) (375.61, 063.75)      /T1_8 in	<|special_separator|>
-(383.93, 056.73) (403.06, 056.73) (403.06, 063.75) (383.93, 063.75)      /T1_8 silico	<|special_separator|>
-(404.66, 056.73) (422.14, 056.73) (422.14, 063.75) (404.66, 063.75)      /T1_8 drug	<|special_separator|>
-(423.74, 056.73) (459.06, 056.73) (459.06, 063.75) (423.74, 063.75)      /T1_8 discovery	<|special_separator|>
-(460.66, 056.73) (467.38, 056.73) (467.38, 063.75) (460.66, 063.75)      /T1_8 in	<|special_separator|>
-(468.98, 056.73) (501.04, 056.73) (501.04, 063.75) (468.98, 063.75)      /T1_8 complex	<|special_separator|>
-(502.64, 056.73) (536.13, 056.73) (536.13, 063.75) (502.64, 063.75)      /T1_8 diseases.	<|special_separator|>
-(537.73, 056.73) (552.32, 056.73) (552.32, 063.75) (537.73, 063.75)      /T1_9 Nat.	<|special_separator|>
-(323.15, 045.98) (352.85, 045.98) (352.85, 052.99) (323.15, 052.99)      /T1_9 Biomed.	<|special_separator|>
-(354.37, 045.98) (370.49, 045.98) (370.49, 052.99) (354.37, 052.99)      /T1_9 Eng.	<|special_separator|>
-(372.07, 046.00) (377.31, 046.00) (377.31, 052.99) (372.07, 052.99)      /T1_6 9	<|special_separator|>
-(377.17, 045.98) (379.43, 045.98) (379.43, 052.99) (377.17, 052.99)      /T1_8 ,	<|special_separator|>
-(381.03, 045.98) (419.08, 045.98) (419.08, 052.99) (381.03, 052.99)      /T1_8 2155-2180	<|special_separator|>
-(420.68, 045.98) (446.52, 045.98) (446.52, 052.99) (420.68, 052.99)      /T1_8 (2025).

groundtruth/parser/font_11.pdf.page_no_12.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2d5e1459a139d7a299f24305f257ee7af848511f93a50bc8838b62ddf887cd33
-size 5455161

groundtruth/parser/font_11.pdf.page_no_12.py.json.line.txt

@@ -1,265 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence | Volume 8 | June 2026 | 984-996	<|special_separator|>
-(547.11, 019.82) (562.68, 019.82) (562.68, 026.80) (547.11, 026.80)      /T1_1 995	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 734.06) (051.32, 734.06) (051.32, 741.07) (039.69, 741.07)      /T1_2 15. 	<|special_separator|>
-(056.69, 734.06) (296.49, 734.06) (296.49, 741.07) (056.69, 741.07)      /T1_2 Qi, X. et al. Predicting transcriptional responses to novel chemical 	<|special_separator|>
-(056.69, 723.30) (285.49, 723.30) (285.49, 730.32) (056.69, 730.32)      /T1_2 perturbations using deep generative model for drug discovery. 	<|special_separator|>
-(056.69, 712.55) (108.65, 712.55) (108.65, 719.57) (056.69, 719.57)      /T1_3 Nat. Commun.	<|special_separator|>
-(108.50, 712.55) (110.62, 712.55) (110.62, 719.57) (108.50, 719.57)      /T1_2  	<|special_separator|>
-(110.22, 712.57) (118.45, 712.57) (118.45, 719.57) (110.22, 719.57)      /T1_4 15	<|special_separator|>
-(118.32, 712.55) (168.22, 712.55) (168.22, 719.57) (118.32, 719.57)      /T1_2 , 9256 (2024).	<|special_separator|>
-(039.69, 701.80) (051.70, 701.80) (051.70, 708.81) (039.69, 708.81)      /T1_2 16. 	<|special_separator|>
-(056.69, 701.80) (265.36, 701.80) (265.36, 708.81) (056.69, 708.81)      /T1_2 He, S. et al. Squidiff: predicting cellular development and 	<|special_separator|>
-(056.69, 691.05) (244.63, 691.05) (244.63, 698.06) (056.69, 698.06)      /T1_2 responses to perturbations using a diffusion model. 	<|special_separator|>
-(244.21, 691.05) (291.73, 691.05) (291.73, 698.06) (244.21, 698.06)      /T1_3 Nat. Methods	<|special_separator|>
-(291.84, 691.05) (293.96, 691.05) (293.96, 698.06) (291.84, 698.06)      /T1_2  	<|special_separator|>
-(056.69, 680.31) (066.38, 680.31) (066.38, 687.31) (056.69, 687.31)      /T1_4 23	<|special_separator|>
-(066.25, 680.30) (119.00, 680.30) (119.00, 687.31) (066.25, 687.31)      /T1_2 , 65-77 (2026).	<|special_separator|>
-(039.69, 669.54) (050.34, 669.54) (050.34, 676.56) (039.69, 676.56)      /T1_2 17. 	<|special_separator|>
-(056.69, 669.54) (272.04, 669.54) (272.04, 676.56) (056.69, 676.56)      /T1_2 Cui, H. et al. scGPT: toward building a foundation model for 	<|special_separator|>
-(056.69, 658.79) (215.03, 658.79) (215.03, 665.81) (056.69, 665.81)      /T1_2 single-cell multi-omics using generative AI. 	<|special_separator|>
-(214.62, 658.79) (262.14, 658.79) (262.14, 665.81) (214.62, 665.81)      /T1_3 Nat. Methods	<|special_separator|>
-(262.00, 658.79) (264.12, 658.79) (264.12, 665.81) (262.00, 665.81)      /T1_2  	<|special_separator|>
-(263.81, 658.81) (271.87, 658.81) (271.87, 665.81) (263.81, 665.81)      /T1_4 21	<|special_separator|>
-(271.73, 658.79) (276.11, 658.79) (276.11, 665.81) (271.73, 665.81)      /T1_2 , 	<|special_separator|>
-(056.69, 648.04) (122.85, 648.04) (122.85, 655.05) (056.69, 655.05)      /T1_2 1470-1480 (2024).	<|special_separator|>
-(039.69, 637.29) (051.60, 637.29) (051.60, 644.30) (039.69, 644.30)      /T1_2 18. 	<|special_separator|>
-(056.69, 637.29) (287.63, 637.29) (287.63, 644.30) (056.69, 644.30)      /T1_2 Yang, F. et al. scBERT as a large-scale pretrained deep language 	<|special_separator|>
-(056.69, 626.54) (269.88, 626.54) (269.88, 633.55) (056.69, 633.55)      /T1_2 model for cell type annotation of single-cell RNA-seq data. 	<|special_separator|>
-(269.46, 626.54) (286.17, 626.54) (286.17, 633.55) (269.46, 633.55)      /T1_3 Nat. 	<|special_separator|>
-(056.69, 615.78) (100.30, 615.78) (100.30, 622.80) (056.69, 622.80)      /T1_3 Mach. Intell.	<|special_separator|>
-(100.16, 615.78) (102.28, 615.78) (102.28, 622.80) (100.16, 622.80)      /T1_2  	<|special_separator|>
-(101.88, 615.80) (107.12, 615.80) (107.12, 622.80) (101.88, 622.80)      /T1_4 4	<|special_separator|>
-(106.99, 615.78) (170.04, 615.78) (170.04, 622.80) (106.99, 622.80)      /T1_2 , 852-866 (2022).	<|special_separator|>
-(039.68, 605.03) (051.39, 605.03) (051.39, 612.05) (039.68, 612.05)      /T1_2 19. 	<|special_separator|>
-(056.69, 605.03) (289.29, 605.03) (289.29, 612.05) (056.69, 612.05)      /T1_2 Noutahi, E. et al. Virtual cells: predict, explain, discover. Preprint 	<|special_separator|>
-(056.69, 594.28) (205.35, 594.28) (205.35, 601.29) (056.69, 601.29)      /T1_2 at http://arxiv.org/abs/2505.14613 (2025).	<|special_separator|>
-(039.68, 583.53) (053.19, 583.53) (053.19, 590.54) (039.68, 590.54)      /T1_2 20. 	<|special_separator|>
-(056.69, 583.53) (276.95, 583.53) (276.95, 590.54) (056.69, 590.54)      /T1_2 Csendes, G., Sanz, G., Szalay, K. Z. & Szalai, B. Benchmarking 	<|special_separator|>
-(056.69, 572.78) (295.04, 572.78) (295.04, 579.79) (056.69, 579.79)      /T1_2 foundation cell models for post-perturbation RNA-seq prediction. 	<|special_separator|>
-(056.69, 562.02) (104.36, 562.02) (104.36, 569.04) (056.69, 569.04)      /T1_3 BMC Genom.	<|special_separator|>
-(104.22, 562.02) (160.45, 562.02) (160.45, 569.04) (104.22, 569.04)      /T1_2  26, 393 (2025).	<|special_separator|>
-(039.68, 551.27) (051.22, 551.27) (051.22, 558.29) (039.68, 558.29)      /T1_2 21. 	<|special_separator|>
-(056.69, 551.27) (285.99, 551.27) (285.99, 558.29) (056.69, 558.29)      /T1_2 Ahlmann-Eltze, C., Huber, W. & Anders, S. Deep-learning-based 	<|special_separator|>
-(056.69, 540.52) (276.39, 540.52) (276.39, 547.53) (056.69, 547.53)      /T1_2 gene perturbation effect prediction does not yet outperform 	<|special_separator|>
-(056.69, 529.77) (143.65, 529.77) (143.65, 536.78) (056.69, 536.78)      /T1_2 simple linear baselines. 	<|special_separator|>
-(143.23, 529.77) (190.76, 529.77) (190.76, 536.78) (143.23, 536.78)      /T1_3 Nat. Methods	<|special_separator|>
-(190.61, 529.77) (268.03, 529.77) (268.03, 536.78) (190.61, 536.78)      /T1_2  22, 1657-1661 (2025).	<|special_separator|>
-(039.68, 519.02) (052.54, 519.02) (052.54, 526.03) (039.68, 526.03)      /T1_2 22. 	<|special_separator|>
-(056.69, 519.02) (293.27, 519.02) (293.27, 526.03) (056.69, 526.03)      /T1_2 Radig, J. et al. scArchon: A scalable benchmarking framework for 	<|special_separator|>
-(056.69, 508.26) (212.16, 508.26) (212.16, 515.28) (056.69, 515.28)      /T1_2 assessing single-cell perturbation models. 	<|special_separator|>
-(211.75, 508.26) (260.01, 508.26) (260.01, 515.28) (211.75, 515.28)      /T1_3 Genome Biol.	<|special_separator|>
-(259.86, 508.26) (288.34, 508.26) (288.34, 515.28) (259.86, 515.28)      /T1_2  https://	<|special_separator|>
-(056.69, 497.51) (216.18, 497.51) (216.18, 504.53) (056.69, 504.53)      /T1_2 doi.org/10.1186/s13059-026-04104-z (2026).	<|special_separator|>
-(039.68, 486.76) (052.86, 486.76) (052.86, 493.77) (039.68, 493.77)      /T1_2 23. 	<|special_separator|>
-(056.69, 486.76) (275.14, 486.76) (275.14, 493.77) (056.69, 493.77)      /T1_2 Maleki, S. et al. Efficient fine-tuning of single-cell foundation 	<|special_separator|>
-(056.69, 476.01) (280.02, 476.01) (280.02, 483.02) (056.69, 483.02)      /T1_2 models enables zero-shot molecular perturbation prediction. 	<|special_separator|>
-(056.69, 465.26) (234.32, 465.26) (234.32, 472.27) (056.69, 472.27)      /T1_2 Preprint at http://arxiv.org/abs/2412.13478 (2025).	<|special_separator|>
-(039.68, 454.50) (052.76, 454.50) (052.76, 461.52) (039.68, 461.52)      /T1_2 24. 	<|special_separator|>
-(056.69, 454.50) (266.58, 454.50) (266.58, 461.52) (056.69, 461.52)      /T1_2 Klein, D., Uscidda, T., Theis, F. & Cuturi, M. Genot: entropic 	<|special_separator|>
-(056.69, 443.75) (256.27, 443.75) (256.27, 450.77) (056.69, 450.77)      /T1_2 (Gromov) Wasserstein flow matching with applications 	<|special_separator|>
-(056.69, 433.00) (154.28, 433.00) (154.28, 440.01) (056.69, 440.01)      /T1_2 to single-cell genomics. In 	<|special_separator|>
-(153.87, 433.00) (252.45, 433.00) (252.45, 440.01) (153.87, 440.01)      /T1_3 38th Conference on Neural 	<|special_separator|>
-(056.69, 422.25) (225.91, 422.25) (225.91, 429.26) (056.69, 429.26)      /T1_3 Information Processing Systems (NeurIPS 2024)	<|special_separator|>
-(225.76, 422.25) (254.19, 422.25) (254.19, 429.26) (225.76, 429.26)      /T1_2  https://	<|special_separator|>
-(056.69, 411.50) (247.18, 411.50) (247.18, 418.51) (056.69, 418.51)      /T1_2 proceedings.neurips.cc/paper_files/paper/2024/file/	<|special_separator|>
-(056.69, 400.74) (280.27, 400.74) (280.27, 407.76) (056.69, 407.76)      /T1_2 bc46e29f91e676747c584ca181cb0ea1-Paper-Conference.pdf 	<|special_separator|>
-(056.69, 389.99) (082.49, 389.99) (082.49, 397.01) (056.69, 397.01)      /T1_2 (2024).	<|special_separator|>
-(039.68, 379.24) (052.56, 379.24) (052.56, 386.25) (039.68, 386.25)      /T1_2 25. 	<|special_separator|>
-(056.69, 379.24) (274.42, 379.24) (274.42, 386.25) (056.69, 386.25)      /T1_2 Bunne, C., Schiebinger, G., Krause, A., Regev, A. & Cuturi, M. 	<|special_separator|>
-(056.69, 368.49) (242.17, 368.49) (242.17, 375.50) (056.69, 375.50)      /T1_2 Optimal transport for single-cell and spatial omics. 	<|special_separator|>
-(241.75, 368.49) (274.73, 368.49) (274.73, 375.50) (241.75, 375.50)      /T1_3 Nat. Rev. 	<|special_separator|>
-(056.69, 357.74) (116.88, 357.74) (116.88, 364.75) (056.69, 364.75)      /T1_3 Methods Primers	<|special_separator|>
-(116.73, 357.74) (118.85, 357.74) (118.85, 364.75) (116.73, 364.75)      /T1_2  	<|special_separator|>
-(118.47, 357.75) (123.72, 357.75) (123.72, 364.75) (118.47, 364.75)      /T1_4 4	<|special_separator|>
-(123.58, 357.74) (164.25, 357.74) (164.25, 364.75) (123.58, 364.75)      /T1_2 , 58 (2024).	<|special_separator|>
-(039.68, 346.98) (052.86, 346.98) (052.86, 354.00) (039.68, 354.00)      /T1_2 26. 	<|special_separator|>
-(056.69, 346.98) (280.19, 346.98) (280.19, 354.00) (056.69, 354.00)      /T1_2 Schiebinger, G. et al. Optimal-transport analysis of single-cell 	<|special_separator|>
-(056.69, 336.23) (261.69, 336.23) (261.69, 343.25) (056.69, 343.25)      /T1_2 gene expression identifies developmental trajectories in 	<|special_separator|>
-(056.69, 325.48) (118.49, 325.48) (118.49, 332.49) (056.69, 332.49)      /T1_2 reprogramming. 	<|special_separator|>
-(118.10, 325.48) (132.55, 325.48) (132.55, 332.49) (118.10, 332.49)      /T1_3 Cell	<|special_separator|>
-(132.41, 325.48) (134.53, 325.48) (134.53, 332.49) (132.41, 332.49)      /T1_2  	<|special_separator|>
-(134.12, 325.50) (146.79, 325.50) (146.79, 332.49) (134.12, 332.49)      /T1_4 176	<|special_separator|>
-(146.64, 325.48) (209.17, 325.48) (209.17, 332.49) (146.64, 332.49)      /T1_2 , 928-943 (2019).	<|special_separator|>
-(039.68, 314.73) (051.50, 314.73) (051.50, 321.74) (039.68, 321.74)      /T1_2 27. 	<|special_separator|>
-(056.68, 314.73) (296.61, 314.73) (296.61, 321.74) (056.68, 321.74)      /T1_2 Klein, D. et al. Mapping cells through time and space with moscot. 	<|special_separator|>
-(056.68, 303.98) (080.81, 303.98) (080.81, 310.99) (056.68, 310.99)      /T1_3 Nature	<|special_separator|>
-(080.67, 303.98) (082.79, 303.98) (082.79, 310.99) (080.67, 310.99)      /T1_2  	<|special_separator|>
-(082.38, 303.99) (097.53, 303.99) (097.53, 310.99) (082.38, 310.99)      /T1_4 638	<|special_separator|>
-(097.40, 303.98) (167.42, 303.98) (167.42, 310.99) (097.40, 310.99)      /T1_2 , 1065-1075 (2025).	<|special_separator|>
-(039.68, 293.22) (052.75, 293.22) (052.75, 300.24) (039.68, 300.24)      /T1_2 28. 	<|special_separator|>
-(056.69, 293.22) (268.82, 293.22) (268.82, 300.24) (056.69, 300.24)      /T1_2 Gossi, F. et al. Matching single cells across modalities with 	<|special_separator|>
-(056.69, 282.47) (213.71, 282.47) (213.71, 289.49) (056.69, 289.49)      /T1_2 contrastive learning and optimal transport. 	<|special_separator|>
-(213.30, 282.47) (270.03, 282.47) (270.03, 289.49) (213.30, 289.49)      /T1_3 Brief. Bioinform.	<|special_separator|>
-(269.89, 282.47) (272.01, 282.47) (272.01, 289.49) (269.89, 289.49)      /T1_2  	<|special_separator|>
-(271.70, 282.49) (281.40, 282.49) (281.40, 289.49) (271.70, 289.49)      /T1_4 24	<|special_separator|>
-(281.26, 282.47) (285.64, 282.47) (285.64, 289.49) (281.26, 289.49)      /T1_2 , 	<|special_separator|>
-(056.69, 271.72) (116.47, 271.72) (116.47, 278.73) (056.69, 278.73)      /T1_2 bbad130 (2023).	<|special_separator|>
-(039.68, 260.97) (052.70, 260.97) (052.70, 267.98) (039.68, 267.98)      /T1_2 29. 	<|special_separator|>
-(056.69, 260.97) (278.33, 260.97) (278.33, 267.98) (056.69, 267.98)      /T1_2 Cao, K., Gong, Q., Hong, Y. & Wan, L. A unified computational 	<|special_separator|>
-(056.69, 250.22) (293.22, 250.22) (293.22, 257.23) (056.69, 257.23)      /T1_2 framework for single-cell data integration with optimal transport. 	<|special_separator|>
-(056.69, 239.46) (108.64, 239.46) (108.64, 246.48) (056.69, 246.48)      /T1_3 Nat. Commun.	<|special_separator|>
-(108.49, 239.46) (110.61, 239.46) (110.61, 246.48) (108.49, 246.48)      /T1_2  	<|special_separator|>
-(110.21, 239.48) (118.61, 239.48) (118.61, 246.48) (110.21, 246.48)      /T1_4 13	<|special_separator|>
-(118.47, 239.46) (165.76, 239.46) (165.76, 246.48) (118.47, 246.48)      /T1_2 , 7419 (2022).	<|special_separator|>
-(039.68, 228.71) (294.19, 228.71) (294.19, 235.73) (039.68, 235.73)      /T1_2 30.  Bunne, C. et al. Learning single-cell perturbation responses using 	<|special_separator|>
-(056.69, 217.96) (148.96, 217.96) (148.96, 224.97) (056.69, 224.97)      /T1_2 neural optimal transport. 	<|special_separator|>
-(148.54, 217.96) (196.07, 217.96) (196.07, 224.97) (148.54, 224.97)      /T1_3 Nat. Methods	<|special_separator|>
-(195.92, 217.96) (198.04, 217.96) (198.04, 224.97) (195.92, 224.97)      /T1_2  	<|special_separator|>
-(197.69, 217.98) (207.96, 217.98) (207.96, 224.97) (197.69, 224.97)      /T1_4 20	<|special_separator|>
-(207.83, 217.96) (276.45, 217.96) (276.45, 224.97) (207.83, 224.97)      /T1_2 , 1759-1768 (2023).	<|special_separator|>
-(039.68, 207.21) (051.58, 207.21) (051.58, 214.22) (039.68, 214.22)      /T1_2 31. 	<|special_separator|>
-(056.69, 207.21) (268.76, 207.21) (268.76, 214.22) (056.69, 214.22)      /T1_2 Jiang, Q., Chen, S., Chen, X. & Jiang, R. scPRAM accurately 	<|special_separator|>
-(056.69, 196.46) (296.37, 196.46) (296.37, 203.47) (056.69, 203.47)      /T1_2 predicts single-cell gene expression perturbation response based 	<|special_separator|>
-(056.69, 185.70) (149.16, 185.70) (149.16, 192.72) (056.69, 192.72)      /T1_2 on attention mechanism. 	<|special_separator|>
-(148.75, 185.70) (200.24, 185.70) (200.24, 192.72) (148.75, 192.72)      /T1_3 Bioinformatics	<|special_separator|>
-(200.09, 185.70) (202.21, 185.70) (202.21, 192.72) (200.09, 192.72)      /T1_2  	<|special_separator|>
-(201.87, 185.72) (212.64, 185.72) (212.64, 192.72) (201.87, 192.72)      /T1_4 40	<|special_separator|>
-(212.51, 185.70) (273.91, 185.70) (273.91, 192.72) (212.51, 192.72)      /T1_2 , btae265 (2024).	<|special_separator|>
-(039.68, 174.95) (052.94, 174.95) (052.94, 181.97) (039.68, 181.97)      /T1_2 32. 	<|special_separator|>
-(056.69, 174.95) (279.77, 174.95) (279.77, 181.97) (056.69, 181.97)      /T1_2 Uscidda, T. & Cuturi, M. The Monge Gap: a regularizer to learn 	<|special_separator|>
-(056.69, 164.20) (135.16, 164.20) (135.16, 171.21) (056.69, 171.21)      /T1_2 all transport maps. In 	<|special_separator|>
-(134.75, 164.20) (275.50, 164.20) (275.50, 171.21) (134.75, 171.21)      /T1_3 Proc. 40th International Conference on 	<|special_separator|>
-(056.69, 153.45) (120.28, 153.45) (120.28, 160.46) (056.69, 160.46)      /T1_3 Machine Learning	<|special_separator|>
-(120.14, 153.45) (277.51, 153.45) (277.51, 160.46) (120.14, 160.46)      /T1_2  (eds Krause, A. et al.) 34709-34733 (PMLR, 	<|special_separator|>
-(056.69, 142.70) (080.11, 142.70) (080.11, 149.71) (056.69, 149.71)      /T1_2 2023).	<|special_separator|>
-(039.68, 131.94) (259.26, 131.94) (259.26, 138.96) (039.68, 138.96)      /T1_2 33.  Bunne, C., Krause, A. & Cuturi, M. Supervised training of 	<|special_separator|>
-(056.69, 121.19) (164.08, 121.19) (164.08, 128.21) (056.69, 128.21)      /T1_2 conditional Monge maps. In 3	<|special_separator|>
-(163.97, 121.19) (258.04, 121.19) (258.04, 128.21) (163.97, 128.21)      /T1_3 6th Conference on Neural 	<|special_separator|>
-(056.69, 110.44) (225.84, 110.44) (225.84, 117.45) (056.69, 117.45)      /T1_3 Information Processing Systems (NeurIPS 2022)	<|special_separator|>
-(225.70, 110.44) (254.13, 110.44) (254.13, 117.45) (225.70, 117.45)      /T1_2  https://	<|special_separator|>
-(056.69, 099.69) (219.46, 099.69) (219.46, 106.70) (056.69, 106.70)      /T1_2 openreview.net/pdf?id=sPNtVVUq7wi (2022).	<|special_separator|>
-(039.68, 088.94) (264.86, 088.94) (264.86, 095.95) (039.68, 095.95)      /T1_2 34.  Vedula, S. et al. Continuous vector quantile regression. In 	<|special_separator|>
-(264.45, 088.94) (285.32, 088.94) (285.32, 095.95) (264.45, 095.95)      /T1_3 ICML 	<|special_separator|>
-(056.69, 078.18) (284.82, 078.18) (284.82, 085.20) (056.69, 085.20)      /T1_3 Workshop on New Frontiers in Learning, Control, and Dynamical 	<|special_separator|>
-(056.69, 067.43) (086.80, 067.43) (086.80, 074.45) (056.69, 074.45)      /T1_3 Systems	<|special_separator|>
-(086.65, 067.43) (276.89, 067.43) (276.89, 074.45) (086.65, 074.45)      /T1_2  https://openreview.net/pdf?id=DUZbGAXcyL (2023).	<|special_separator|>
-(039.68, 056.68) (052.96, 056.68) (052.96, 063.69) (039.68, 063.69)      /T1_2 35. 	<|special_separator|>
-(056.69, 056.68) (283.45, 056.68) (283.45, 063.69) (056.69, 063.69)      /T1_2 Yang, K. et al. Analyzing learned molecular representations for 	<|special_separator|>
-(056.69, 045.93) (131.78, 045.93) (131.78, 052.94) (056.69, 052.94)      /T1_2 property prediction. 	<|special_separator|>
-(131.37, 045.93) (200.96, 045.93) (200.96, 052.94) (131.37, 052.94)      /T1_3 J. Chem. Inf. Model.	<|special_separator|>
-(200.82, 045.93) (202.94, 045.93) (202.94, 052.94) (200.82, 052.94)      /T1_2  	<|special_separator|>
-(202.58, 045.94) (212.52, 045.94) (212.52, 052.94) (202.58, 052.94)      /T1_4 59	<|special_separator|>
-(212.39, 045.93) (284.03, 045.93) (284.03, 052.94) (212.39, 052.94)      /T1_2 , 3370-3388 (2019).	<|special_separator|>
-(306.15, 733.93) (539.42, 733.93) (539.42, 740.94) (306.15, 740.94)      /T1_2 36.  Gut, G., Herrmann, M. D. & Pelkmans, L. Multiplexed protein 	<|special_separator|>
-(323.16, 723.18) (513.14, 723.18) (513.14, 730.19) (323.16, 730.19)      /T1_2 maps link subcellular organization to cellular states. 	<|special_separator|>
-(512.73, 723.18) (541.37, 723.18) (541.37, 730.19) (512.73, 730.19)      /T1_3 Science	<|special_separator|>
-(541.22, 723.18) (543.34, 723.18) (543.34, 730.19) (541.22, 730.19)      /T1_2  	<|special_separator|>
-(543.04, 723.19) (556.49, 723.19) (556.49, 730.19) (543.04, 730.19)      /T1_4 361	<|special_separator|>
-(556.35, 723.18) (560.72, 723.18) (560.72, 730.19) (556.35, 730.19)      /T1_2 , 	<|special_separator|>
-(323.16, 712.42) (383.65, 712.42) (383.65, 719.44) (323.16, 719.44)      /T1_2 eaar7042 (2018).	<|special_separator|>
-(306.15, 701.67) (318.25, 701.67) (318.25, 708.69) (306.15, 708.69)      /T1_2 37. 	<|special_separator|>
-(323.16, 701.67) (533.04, 701.67) (533.04, 708.69) (323.16, 708.69)      /T1_2 Hao, M. et al. Large-scale foundation model on single-cell 	<|special_separator|>
-(323.16, 690.92) (384.64, 690.92) (384.64, 697.93) (323.16, 697.93)      /T1_2 transcriptomics. 	<|special_separator|>
-(384.26, 690.92) (431.78, 690.92) (431.78, 697.93) (384.26, 697.93)      /T1_3 Nat. Methods	<|special_separator|>
-(431.64, 690.92) (433.76, 690.92) (433.76, 697.93) (431.64, 697.93)      /T1_2  	<|special_separator|>
-(433.36, 690.94) (441.42, 690.94) (441.42, 697.93) (433.36, 697.93)      /T1_4 21	<|special_separator|>
-(441.28, 690.92) (507.90, 690.92) (507.90, 697.93) (441.28, 697.93)      /T1_2 , 1481-1491 (2024).	<|special_separator|>
-(306.15, 680.17) (319.46, 680.17) (319.46, 687.18) (306.15, 687.18)      /T1_2 38. 	<|special_separator|>
-(323.16, 680.17) (554.95, 680.17) (554.95, 687.18) (323.16, 687.18)      /T1_2 Viñas Torné, R. et al. Systema: a framework for evaluating genetic 	<|special_separator|>
-(323.16, 669.42) (546.22, 669.42) (546.22, 676.43) (323.16, 676.43)      /T1_2 perturbation response prediction beyond systematic variation. 	<|special_separator|>
-(545.73, 669.42) (562.20, 669.42) (562.20, 676.43) (545.73, 676.43)      /T1_3 Nat. 	<|special_separator|>
-(323.16, 658.66) (360.87, 658.66) (360.87, 665.68) (323.16, 665.68)      /T1_3 Biotechnol	<|special_separator|>
-(360.65, 658.66) (545.77, 658.66) (545.77, 665.68) (360.65, 665.68)      /T1_2 . https://doi.org/10.1038/s41587-025-02777-8 (2025).	<|special_separator|>
-(306.15, 647.91) (319.21, 647.91) (319.21, 654.93) (306.15, 654.93)      /T1_2 39. 	<|special_separator|>
-(323.16, 647.91) (549.77, 647.91) (549.77, 654.93) (323.16, 654.93)      /T1_2 Driessen, A., Born, J., Rueda, R. C., Reddy, S. T . & Rapsomaniki, M. 	<|special_separator|>
-(323.16, 637.16) (545.76, 637.16) (545.76, 644.17) (323.16, 644.17)      /T1_2 Modeling chimeric antigen receptor response at the single-cell 	<|special_separator|>
-(323.16, 626.41) (465.05, 626.41) (465.05, 633.42) (323.16, 633.42)      /T1_2 level with conditional optimal transport. 	<|special_separator|>
-(464.51, 626.41) (509.27, 626.41) (509.27, 633.42) (464.51, 633.42)      /T1_3 Cell Systems	<|special_separator|>
-(509.01, 626.41) (560.73, 626.41) (560.73, 633.42) (509.01, 633.42)      /T1_2  101591 (2026).	<|special_separator|>
-(306.15, 615.66) (528.55, 615.66) (528.55, 622.67) (306.15, 622.67)      /T1_2 40.  Lübeck, F. et al. Neural unbalanced optimal transport via 	<|special_separator|>
-(323.16, 604.90) (540.93, 604.90) (540.93, 611.92) (323.16, 611.92)      /T1_2 cycle-consistent semi-couplings. Preprint at http://arxiv.org/	<|special_separator|>
-(323.16, 594.15) (405.87, 594.15) (405.87, 601.17) (323.16, 601.17)      /T1_2 abs/2209.15621 (2022).	<|special_separator|>
-(306.15, 583.40) (318.11, 583.40) (318.11, 590.41) (306.15, 590.41)      /T1_2 41. 	<|special_separator|>
-(323.16, 583.40) (522.45, 583.40) (522.45, 590.41) (323.16, 590.41)      /T1_2 Eyring, L. et al. Unbalancedness in neural Monge maps 	<|special_separator|>
-(323.16, 572.65) (473.68, 572.65) (473.68, 579.66) (323.16, 579.66)      /T1_2 improves unpaired domain translation. In 	<|special_separator|>
-(473.26, 572.65) (556.76, 572.65) (556.76, 579.66) (473.26, 579.66)      /T1_3 Proc. 12th International 	<|special_separator|>
-(323.16, 561.90) (468.46, 561.90) (468.46, 568.91) (323.16, 568.91)      /T1_3 Conference on Learning Representations	<|special_separator|>
-(468.31, 561.90) (555.45, 561.90) (555.45, 568.91) (468.31, 568.91)      /T1_2  https://openreview.net/	<|special_separator|>
-(323.16, 551.14) (419.93, 551.14) (419.93, 558.16) (323.16, 558.16)      /T1_2 pdf?id=2UnCj3jeao (2024).	<|special_separator|>
-(306.15, 540.39) (319.50, 540.39) (319.50, 547.41) (306.15, 547.41)      /T1_2 42. 	<|special_separator|>
-(323.16, 540.39) (557.89, 540.39) (557.89, 547.41) (323.16, 547.41)      /T1_2 Zhang, J. et al. Tahoe-100M: a giga-scale single-cell perturbation 	<|special_separator|>
-(323.16, 529.64) (562.42, 529.64) (562.42, 536.65) (323.16, 536.65)      /T1_2 atlas for context-dependent gene function and cellular modeling. 	<|special_separator|>
-(323.16, 518.89) (362.87, 518.89) (362.87, 525.90) (323.16, 525.90)      /T1_2 Preprint at 	<|special_separator|>
-(362.46, 518.89) (388.47, 518.89) (388.47, 525.90) (362.46, 525.90)      /T1_3 bioRxiv	<|special_separator|>
-(388.33, 518.89) (545.62, 518.89) (545.62, 525.90) (388.33, 525.90)      /T1_2  https://doi.org/10.1101/2025.02.20.639398 	<|special_separator|>
-(323.16, 508.14) (348.99, 508.14) (348.99, 515.15) (323.16, 515.15)      /T1_2 (2025).	<|special_separator|>
-(306.15, 497.38) (552.26, 497.38) (552.26, 504.40) (306.15, 504.40)      /T1_2 43.  Borgwardt, K. M. et al. Integrating structured biological data by 	<|special_separator|>
-(323.16, 486.63) (456.18, 486.63) (456.18, 493.65) (323.16, 493.65)      /T1_2 kernel maximum mean discrepancy. 	<|special_separator|>
-(455.76, 486.63) (507.25, 486.63) (507.25, 493.65) (455.76, 493.65)      /T1_3 Bioinformatics	<|special_separator|>
-(507.11, 486.63) (509.23, 486.63) (509.23, 493.65) (507.11, 493.65)      /T1_2  	<|special_separator|>
-(508.89, 486.65) (518.28, 486.65) (518.28, 493.65) (508.89, 493.65)      /T1_4 22	<|special_separator|>
-(518.15, 486.63) (555.31, 486.63) (555.31, 493.65) (518.15, 493.65)      /T1_2 , e49-e57 	<|special_separator|>
-(323.16, 475.88) (350.27, 475.88) (350.27, 482.89) (323.16, 482.89)      /T1_2 (2006).	<|special_separator|>
-(306.16, 465.13) (532.31, 465.13) (532.31, 472.14) (306.16, 472.14)      /T1_2 44.  Cuturi, M. Sinkhorn distances: lightspeed computation of 	<|special_separator|>
-(323.16, 454.38) (399.14, 454.38) (399.14, 461.39) (323.16, 461.39)      /T1_2 optimal transport. In 	<|special_separator|>
-(398.72, 454.38) (526.77, 454.38) (526.77, 461.39) (398.72, 461.39)      /T1_3 Proc. 27th International Conference 	<|special_separator|>
-(323.16, 443.62) (473.20, 443.62) (473.20, 450.64) (323.16, 450.64)      /T1_3 on Neural Information Processing Systems	<|special_separator|>
-(473.06, 443.62) (501.48, 443.62) (501.48, 450.64) (473.06, 450.64)      /T1_2  https://	<|special_separator|>
-(323.16, 432.87) (512.35, 432.87) (512.35, 439.89) (323.16, 439.89)      /T1_2 proceedings.neurips.cc/paper_files/paper/2013/file/	<|special_separator|>
-(323.16, 422.12) (526.81, 422.12) (526.81, 429.13) (323.16, 429.13)      /T1_2 af21d0c97db2e27e13572cbf59eb343d-Paper.pdf (2013).	<|special_separator|>
-(306.16, 411.37) (319.54, 411.37) (319.54, 418.38) (306.16, 418.38)      /T1_2 45. 	<|special_separator|>
-(323.16, 411.37) (550.89, 411.37) (550.89, 418.38) (323.16, 418.38)      /T1_2 Genevay, A., Peyre, G. & Cuturi, M. Learning generative models 	<|special_separator|>
-(323.16, 400.62) (432.32, 400.62) (432.32, 407.63) (323.16, 407.63)      /T1_2 with Sinkhorn divergences. In 	<|special_separator|>
-(431.90, 400.62) (558.21, 400.62) (558.21, 407.63) (431.90, 407.63)      /T1_3 Proc. 21st International Conference 	<|special_separator|>
-(323.16, 389.86) (456.32, 389.86) (456.32, 396.88) (323.16, 396.88)      /T1_3 on Artificial Intelligence and Statistics	<|special_separator|>
-(456.18, 389.86) (522.88, 389.86) (522.88, 396.88) (456.18, 396.88)      /T1_2  (eds Storkey, A. & 	<|special_separator|>
-(323.16, 379.11) (463.74, 379.11) (463.74, 386.13) (323.16, 386.13)      /T1_2 Perez-Cruz, F.) 1608-1617 (PMLR, 2018).	<|special_separator|>
-(306.16, 368.36) (537.45, 368.36) (537.45, 375.37) (306.16, 375.37)      /T1_2 46.  Feydy, J. et al. Interpolating between optimal transport and 	<|special_separator|>
-(323.16, 357.61) (457.28, 357.61) (457.28, 364.62) (323.16, 364.62)      /T1_2 MMD using Sinkhorn divergences. In 	<|special_separator|>
-(456.87, 357.61) (543.96, 357.61) (543.96, 364.62) (456.87, 364.62)      /T1_3 Proc. 22nd International 	<|special_separator|>
-(323.16, 346.86) (500.03, 346.86) (500.03, 353.87) (323.16, 353.87)      /T1_3 Conference on Artificial Intelligence and Statistics	<|special_separator|>
-(499.88, 346.86) (561.51, 346.86) (561.51, 353.87) (499.88, 353.87)      /T1_2  (eds Chaudhuri, 	<|special_separator|>
-(323.16, 336.10) (482.57, 336.10) (482.57, 343.12) (323.16, 343.12)      /T1_2 K. & Sugiyama, M.) 2681-2690 (PMLR, 2019).	<|special_separator|>
-(306.16, 325.35) (318.40, 325.35) (318.40, 332.37) (306.16, 332.37)      /T1_2 47. 	<|special_separator|>
-(323.16, 325.35) (549.75, 325.35) (549.75, 332.37) (323.16, 332.37)      /T1_2 Brenier, Y. Décomposition polaire et réarrangement monotone 	<|special_separator|>
-(323.16, 314.60) (415.53, 314.60) (415.53, 321.61) (323.16, 321.61)      /T1_2 des champs de vecteurs. 	<|special_separator|>
-(415.10, 314.60) (481.74, 314.60) (481.74, 321.61) (415.10, 321.61)      /T1_3 CR Acad. Sci. Sér. I	<|special_separator|>
-(481.60, 314.60) (483.72, 314.60) (483.72, 321.61) (481.60, 321.61)      /T1_2  	<|special_separator|>
-(483.35, 314.62) (498.78, 314.62) (498.78, 321.61) (483.35, 321.61)      /T1_4 305	<|special_separator|>
-(498.64, 314.60) (561.26, 314.60) (561.26, 321.61) (498.64, 321.61)      /T1_2 , 805-808 (1987).	<|special_separator|>
-(306.16, 303.85) (547.83, 303.85) (547.83, 310.86) (306.16, 310.86)      /T1_2 48.  Amos, B., Xu, L. & Kolter, J. Z. Input convex neural networks. In 	<|special_separator|>
-(323.16, 293.10) (526.16, 293.10) (526.16, 300.11) (323.16, 300.11)      /T1_3 Proc. 34th International Conference on Machine Learning	<|special_separator|>
-(526.02, 293.10) (545.43, 293.10) (545.43, 300.11) (526.02, 300.11)      /T1_2  (eds 	<|special_separator|>
-(323.16, 282.34) (484.62, 282.34) (484.62, 289.36) (323.16, 289.36)      /T1_2 Precup, D. & Teh, Y. W.) 146-155 (PMLR, 2017).	<|special_separator|>
-(306.16, 271.59) (319.68, 271.59) (319.68, 278.61) (306.16, 278.61)      /T1_2 49. 	<|special_separator|>
-(323.16, 271.59) (539.14, 271.59) (539.14, 278.61) (323.16, 278.61)      /T1_2 Makkuva, A., Taghvaei, A., Oh, S. & Lee, J. Optimal transport 	<|special_separator|>
-(323.16, 260.84) (489.14, 260.84) (489.14, 267.85) (323.16, 267.85)      /T1_2 mapping via input convex neural networks. In 	<|special_separator|>
-(488.72, 260.84) (526.72, 260.84) (526.72, 267.85) (488.72, 267.85)      /T1_3 Proc. 37th 	<|special_separator|>
-(323.16, 250.09) (487.92, 250.09) (487.92, 257.10) (323.16, 257.10)      /T1_3 International Conference on Machine Learning	<|special_separator|>
-(487.78, 250.09) (561.67, 250.09) (561.67, 257.10) (487.78, 257.10)      /T1_2  (eds Daumé, H. III & 	<|special_separator|>
-(323.16, 239.34) (451.39, 239.34) (451.39, 246.35) (323.16, 246.35)      /T1_2 Singh, A.) 6672-6681 (PMLR, 2020).	<|special_separator|>
-(306.16, 228.58) (524.54, 228.58) (524.54, 235.60) (306.16, 235.60)      /T1_2 50.  Chen, Y., Hu, Z., Chen, W. & Huang, H. Fast and scalable 	<|special_separator|>
-(323.16, 217.83) (539.39, 217.83) (539.39, 224.85) (323.16, 224.85)      /T1_2 Wasserstein-1 neural optimal transport solver for single-cell 	<|special_separator|>
-(323.16, 207.08) (412.46, 207.08) (412.46, 214.09) (323.16, 214.09)      /T1_2 perturbation prediction. 	<|special_separator|>
-(412.04, 207.08) (463.53, 207.08) (463.53, 214.09) (412.04, 214.09)      /T1_3 Bioinformatics	<|special_separator|>
-(463.39, 207.08) (465.51, 207.08) (465.51, 214.09) (463.39, 214.09)      /T1_2  	<|special_separator|>
-(465.16, 207.10) (473.60, 207.10) (473.60, 214.09) (465.16, 214.09)      /T1_4 41	<|special_separator|>
-(473.47, 207.08) (538.50, 207.08) (538.50, 214.09) (473.47, 214.09)      /T1_2 , i513-i522 (2025).	<|special_separator|>
-(306.16, 196.33) (317.95, 196.33) (317.95, 203.34) (306.16, 203.34)      /T1_2 51. 	<|special_separator|>
-(323.16, 196.33) (540.06, 196.33) (540.06, 203.34) (323.16, 203.34)      /T1_2 Rosenberg, A. A., Vedula, S., Romano, Y. & Bronstein, A. Fast 	<|special_separator|>
-(323.16, 185.58) (466.94, 185.58) (466.94, 192.59) (323.16, 192.59)      /T1_2 nonlinear vector quantile regression. In 	<|special_separator|>
-(466.52, 185.58) (548.78, 185.58) (548.78, 192.59) (466.52, 192.59)      /T1_3 Proc. 11th International 	<|special_separator|>
-(323.16, 174.82) (468.46, 174.82) (468.46, 181.84) (323.16, 181.84)      /T1_3 Conference on Learning Representations	<|special_separator|>
-(468.32, 174.82) (496.00, 174.82) (496.00, 181.84) (468.32, 181.84)      /T1_2  (2023).	<|special_separator|>
-(306.16, 164.07) (319.27, 164.07) (319.27, 171.09) (306.16, 171.09)      /T1_2 52. 	<|special_separator|>
-(323.16, 164.07) (546.85, 164.07) (546.85, 171.09) (323.16, 171.09)      /T1_2 Pegoraro, M. et al. Vector quantile regression on manifolds. In 	<|special_separator|>
-(323.16, 153.32) (551.92, 153.32) (551.92, 160.33) (323.16, 160.33)      /T1_3 Proc. 27th International Conference on Artificial Intelligence and 	<|special_separator|>
-(323.16, 142.57) (355.62, 142.57) (355.62, 149.58) (323.16, 149.58)      /T1_3 Statistics	<|special_separator|>
-(355.48, 142.57) (555.60, 142.57) (555.60, 149.58) (355.48, 149.58)      /T1_2  (eds Dasgupta, S., Mandt, S. & Li, Y.) 1999-2007 (PMLR, 	<|special_separator|>
-(323.16, 131.82) (346.41, 131.82) (346.41, 138.83) (323.16, 138.83)      /T1_2 2024).	<|special_separator|>
-(306.16, 121.06) (557.73, 121.06) (557.73, 128.08) (306.16, 128.08)      /T1_2 53.  Tong, A. et al. Improving and generalizing flow-based generative 	<|special_separator|>
-(323.16, 110.31) (483.81, 110.31) (483.81, 117.33) (323.16, 117.33)      /T1_2 models with minibatch optimal transport. In 	<|special_separator|>
-(483.39, 110.31) (501.94, 110.31) (501.94, 117.33) (483.39, 117.33)      /T1_3 Proc.	<|special_separator|>
-(501.80, 110.31) (503.92, 110.31) (503.92, 117.33) (501.80, 117.33)      /T1_2  	<|special_separator|>
-(503.59, 110.31) (561.92, 110.31) (561.92, 117.33) (503.59, 117.33)      /T1_3 ICML Workshop 	<|special_separator|>
-(323.16, 099.56) (543.34, 099.56) (543.34, 106.57) (323.16, 106.57)      /T1_3 on New Frontiers in Learning, Control, and Dynamical Systems	<|special_separator|>
-(543.42, 099.56) (545.54, 099.56) (545.54, 106.57) (543.42, 106.57)      /T1_2  	<|special_separator|>
-(323.16, 088.81) (513.85, 088.81) (513.85, 095.82) (323.16, 095.82)      /T1_2 https://openreview.net/pdf?id=CD9Snc73AW (2023).	<|special_separator|>
-(306.16, 078.06) (559.95, 078.06) (559.95, 085.07) (306.16, 085.07)      /T1_2 54.  Pooladian, A.-A. et al. Multisample flow matching: straightening 	<|special_separator|>
-(323.16, 067.30) (451.76, 067.30) (451.76, 074.32) (323.16, 074.32)      /T1_2 flows with minibatch couplings. In 	<|special_separator|>
-(451.42, 067.30) (540.40, 067.30) (540.40, 074.32) (451.42, 074.32)      /T1_3 Proc. 40th International 	<|special_separator|>
-(323.16, 056.55) (444.76, 056.55) (444.76, 063.57) (323.16, 063.57)      /T1_3 Conference on Machine Learning	<|special_separator|>
-(444.74, 056.55) (525.03, 056.55) (525.03, 063.57) (444.74, 063.57)      /T1_2  (eds Krause, A. et al)  	<|special_separator|>
-(323.16, 045.80) (424.05, 045.80) (424.05, 052.81) (323.16, 052.81)      /T1_2 28100-28127 (PMLR, 2023).

groundtruth/parser/font_11.pdf.page_no_13.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:996946da528b1206bc058abbc110c838d661fa91a5199a5dc2f5b70fe3369ec7
-size 3516130

groundtruth/parser/font_11.pdf.page_no_13.py.json.line.txt

@@ -1,134 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence | Volume 8 | June 2026 | 984-996	<|special_separator|>
-(546.10, 020.07) (562.05, 020.07) (562.05, 027.05) (546.10, 027.05)      /T1_1 996	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 734.06) (052.75, 734.06) (052.75, 741.07) (039.69, 741.07)      /T1_2 55. 	<|special_separator|>
-(056.69, 734.06) (288.16, 734.06) (288.16, 741.07) (056.69, 741.07)      /T1_2 Mariella, N. et al. Quantum theory and application of contextual 	<|special_separator|>
-(056.69, 723.30) (132.67, 723.30) (132.67, 730.32) (056.69, 730.32)      /T1_2 optimal transport. In 	<|special_separator|>
-(132.25, 723.30) (269.81, 723.30) (269.81, 730.32) (132.25, 730.32)      /T1_3 Proc. 41st International Conference on 	<|special_separator|>
-(056.69, 712.55) (120.28, 712.55) (120.28, 719.57) (056.69, 719.57)      /T1_3 Machine Learning	<|special_separator|>
-(120.14, 712.55) (258.54, 712.55) (258.54, 719.57) (120.14, 719.57)      /T1_2 , Vol. 235, 34822-34845 (PMLR, 2024).	<|special_separator|>
-(039.69, 701.80) (053.12, 701.80) (053.12, 708.81) (039.69, 708.81)      /T1_2 56. 	<|special_separator|>
-(056.69, 701.80) (265.81, 701.80) (265.81, 708.81) (056.69, 708.81)      /T1_2 Basu, S. et al. Advancing single-cell omics and cell-based 	<|special_separator|>
-(056.69, 691.05) (200.68, 691.05) (200.68, 698.06) (056.69, 698.06)      /T1_2 therapeutics with quantum computing. 	<|special_separator|>
-(200.26, 691.05) (281.69, 691.05) (281.69, 698.06) (200.26, 698.06)      /T1_3 Nat. Rev. Mol. Cell Biol.	<|special_separator|>
-(281.54, 691.05) (283.66, 691.05) (283.66, 698.06) (281.54, 698.06)      /T1_2  	<|special_separator|>
-(283.34, 691.06) (292.37, 691.06) (292.37, 698.06) (283.34, 698.06)      /T1_4 27	<|special_separator|>
-(292.24, 691.05) (296.61, 691.05) (296.61, 698.06) (292.24, 698.06)      /T1_2 , 	<|special_separator|>
-(056.69, 680.30) (117.51, 680.30) (117.51, 687.31) (056.69, 687.31)      /T1_2 394-408 (2026).	<|special_separator|>
-(039.68, 669.54) (051.57, 669.54) (051.57, 676.56) (039.68, 676.56)      /T1_2 57. 	<|special_separator|>
-(056.69, 669.54) (293.22, 669.54) (293.22, 676.56) (056.69, 676.56)      /T1_2 Ramdas, A., Trillos, N. G. & Cuturi, M. On Wasserstein two-sample 	<|special_separator|>
-(056.69, 658.79) (243.06, 658.79) (243.06, 665.81) (056.69, 665.81)      /T1_2 testing and related families of nonparametric tests. 	<|special_separator|>
-(242.65, 658.79) (270.40, 658.79) (270.40, 665.81) (242.65, 665.81)      /T1_3 Entropy	<|special_separator|>
-(270.26, 658.79) (272.38, 658.79) (272.38, 665.81) (270.26, 665.81)      /T1_2  	<|special_separator|>
-(272.06, 658.81) (280.57, 658.81) (280.57, 665.81) (272.06, 665.81)      /T1_4 19	<|special_separator|>
-(280.44, 658.79) (286.41, 658.79) (286.41, 665.81) (280.44, 665.81)      /T1_2 ,  	<|special_separator|>
-(056.69, 648.04) (091.35, 648.04) (091.35, 655.05) (056.69, 655.05)      /T1_2 47 (2017).	<|special_separator|>
-(039.68, 637.29) (053.07, 637.29) (053.07, 644.30) (039.68, 644.30)      /T1_2 58. 	<|special_separator|>
-(056.69, 637.29) (295.55, 637.29) (295.55, 644.30) (056.69, 644.30)      /T1_2 Salimans, T., Zhang, H., Radford, A. & Metaxas, D. Improving GANs 	<|special_separator|>
-(056.69, 626.54) (154.28, 626.54) (154.28, 633.55) (056.69, 633.55)      /T1_2 using optimal transport. In 	<|special_separator|>
-(153.87, 626.54) (288.56, 626.54) (288.56, 633.55) (153.87, 633.55)      /T1_3 International Conference on Learning 	<|special_separator|>
-(056.69, 615.78) (114.53, 615.78) (114.53, 622.80) (056.69, 622.80)      /T1_3 Representations	<|special_separator|>
-(114.39, 615.78) (294.60, 615.78) (294.60, 622.80) (114.39, 622.80)      /T1_2  https://openreview.net/pdf?id=rkQkBnJAb (2018).	<|special_separator|>
-(039.68, 605.03) (052.89, 605.03) (052.89, 612.05) (039.68, 612.05)      /T1_2 59. 	<|special_separator|>
-(056.69, 605.03) (283.54, 605.03) (283.54, 612.05) (056.69, 612.05)      /T1_2 Hendrycks, D. & Gimpel, K. Gaussian error linear units (GELUs). 	<|special_separator|>
-(056.69, 594.28) (237.84, 594.28) (237.84, 601.29) (056.69, 601.29)      /T1_2 Preprint at http://arxiv.org/abs/1606.08415 (2023).	<|special_separator|>
-(039.68, 583.53) (285.54, 583.53) (285.54, 590.54) (039.68, 590.54)      /T1_2 60.  de Leeuw, J. Application of convex analysis to multidimensional 	<|special_separator|>
-(056.69, 572.78) (086.82, 572.78) (086.82, 579.79) (056.69, 579.79)      /T1_2 scaling. 	<|special_separator|>
-(086.38, 572.78) (219.02, 572.78) (219.02, 579.79) (086.38, 579.79)      /T1_3 UCLA: Department of Statistics, UCLA	<|special_separator|>
-(218.84, 572.78) (294.51, 572.78) (294.51, 579.79) (218.84, 579.79)      /T1_2  https://escholarship.	<|special_separator|>
-(056.69, 562.02) (163.14, 562.02) (163.14, 569.04) (056.69, 569.04)      /T1_2 org/uc/item/4ps3b5mj (2011).	<|special_separator|>
-(039.68, 551.27) (051.74, 551.27) (051.74, 558.29) (039.68, 558.29)      /T1_2 61. 	<|special_separator|>
-(056.69, 551.27) (163.58, 551.27) (163.58, 558.29) (056.69, 558.29)      /T1_2 Zaheer, M. et al. Deep sets. In 	<|special_separator|>
-(163.16, 551.27) (233.95, 551.27) (233.95, 558.29) (163.16, 558.29)      /T1_3 31st Conference on 	<|special_separator|>
-(056.69, 540.52) (237.92, 540.52) (237.92, 547.53) (056.69, 547.53)      /T1_3 Neural Information Processing Systems (NIPS 2017)	<|special_separator|>
-(237.99, 540.52) (240.11, 540.52) (240.11, 547.53) (237.99, 547.53)      /T1_2  	<|special_separator|>
-(056.69, 529.77) (238.23, 529.77) (238.23, 536.78) (056.69, 536.78)      /T1_2 https://papers.nips.cc/paper_files/paper/2017/file/	<|special_separator|>
-(056.69, 519.02) (256.08, 519.02) (256.08, 526.03) (056.69, 526.03)      /T1_2 f22e4747da1aa27e363d86d40ff442fe-Paper.pdf (2017).	<|special_separator|>
-(039.68, 508.26) (053.06, 508.26) (053.06, 515.28) (039.68, 515.28)      /T1_2 62. 	<|special_separator|>
-(056.69, 508.26) (296.18, 508.26) (296.18, 515.28) (056.69, 515.28)      /T1_2 Wolf, F. A., Angerer, P. & Theis, F. J. SCANPY: large-scale single-cell 	<|special_separator|>
-(056.69, 497.51) (169.00, 497.51) (169.00, 504.53) (056.69, 504.53)      /T1_2 gene expression data analysis. 	<|special_separator|>
-(168.58, 497.51) (216.84, 497.51) (216.84, 504.53) (168.58, 504.53)      /T1_3 Genome Biol.	<|special_separator|>
-(216.70, 497.51) (218.82, 497.51) (218.82, 504.53) (216.70, 504.53)      /T1_2  	<|special_separator|>
-(218.48, 497.53) (226.99, 497.53) (226.99, 504.53) (218.48, 504.53)      /T1_4 19	<|special_separator|>
-(226.85, 497.51) (264.77, 497.51) (264.77, 504.53) (226.85, 504.53)      /T1_2 , 15 (2018).	<|special_separator|>
-(039.69, 486.76) (293.19, 486.76) (293.19, 493.77) (039.69, 493.77)      /T1_2 63.  Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. 	<|special_separator|>
-(056.70, 476.01) (065.61, 476.01) (065.61, 483.02) (056.70, 483.02)      /T1_2 In 	<|special_separator|>
-(065.20, 476.01) (257.13, 476.01) (257.13, 483.02) (065.20, 483.02)      /T1_3 International Conference on Learning Representations	<|special_separator|>
-(256.99, 476.01) (285.43, 476.01) (285.43, 483.02) (256.99, 483.02)      /T1_2  https://	<|special_separator|>
-(056.69, 465.26) (213.91, 465.26) (213.91, 472.27) (056.69, 472.27)      /T1_2 openreview.net/pdf?id=Bkg6RiCqY7 (2019).	<|special_separator|>
-(039.68, 454.50) (052.88, 454.50) (052.88, 461.52) (039.68, 461.52)      /T1_2 64. 	<|special_separator|>
-(056.69, 454.50) (285.34, 454.50) (285.34, 461.52) (056.69, 461.52)      /T1_2 Cuturi, M. et al. Optimal Transport Tools (OTT): a JAX toolbox for all 	<|special_separator|>
-(056.69, 443.75) (292.57, 443.75) (292.57, 450.77) (056.69, 450.77)      /T1_2 things Wasserstein. Preprint at http://arxiv.org/abs/2201.12324 (2022).	<|special_separator|>
-(039.68, 433.00) (053.08, 433.00) (053.08, 440.01) (039.68, 440.01)      /T1_2 65. 	<|special_separator|>
-(056.69, 433.00) (275.17, 433.00) (275.17, 440.01) (056.69, 440.01)      /T1_2 Kingma, D. P. & Welling, M. Auto-encoding variational Bayes. 	<|special_separator|>
-(056.69, 422.25) (230.88, 422.25) (230.88, 429.26) (056.69, 429.26)      /T1_2 Preprint at https://arxiv.org/abs/1312.6114 (2013).	<|special_separator|>
-(039.68, 411.50) (273.86, 411.50) (273.86, 418.51) (039.68, 418.51)      /T1_2 66.  Bunne, C. et al. Datasets of learning single-cell perturbation 	<|special_separator|>
-(056.69, 400.74) (209.57, 400.74) (209.57, 407.76) (056.69, 407.76)      /T1_2 responses using neural optimal transport. 	<|special_separator|>
-(209.22, 400.74) (223.84, 400.74) (223.84, 407.76) (209.22, 407.76)      /T1_3 ETH	<|special_separator|>
-(223.69, 400.74) (225.81, 400.74) (225.81, 407.76) (223.69, 407.76)      /T1_2  	<|special_separator|>
-(225.40, 400.74) (248.19, 400.74) (248.19, 407.76) (225.40, 407.76)      /T1_3 Zürich	<|special_separator|>
-(248.05, 400.74) (289.67, 400.74) (289.67, 407.76) (248.05, 407.76)      /T1_2  https://doi.	<|special_separator|>
-(056.70, 389.99) (200.10, 389.99) (200.10, 397.01) (056.70, 397.01)      /T1_2 org/10.3929/ethz-b-000609681 (2023).	<|special_separator|>
-(039.69, 379.24) (051.95, 379.24) (051.95, 386.25) (039.69, 386.25)      /T1_2 67. 	<|special_separator|>
-(056.70, 379.24) (287.28, 379.24) (287.28, 386.25) (056.70, 386.25)      /T1_2 Born, J., Driessen, A. & Harsanyi, B. AI4SCR/conditional-monge- 	<|special_separator|>
-(056.70, 368.49) (074.86, 368.49) (074.86, 375.50) (056.70, 375.50)      /T1_2 gap. 	<|special_separator|>
-(074.44, 368.49) (101.93, 368.49) (101.93, 375.50) (074.44, 375.50)      /T1_3 Zenodo	<|special_separator|>
-(101.78, 368.49) (279.38, 368.49) (279.38, 375.50) (101.78, 375.50)      /T1_2  https://doi.org/10.5281/zenodo.17854338 (2025).	<|special_separator|>
-(039.69, 346.35) (140.85, 346.35) (140.85, 356.11) (039.69, 356.11)      /T1_5 Acknowledgements	<|special_separator|>
-(039.69, 336.31) (278.00, 336.31) (278.00, 343.32) (039.69, 343.32)      /T1_2 We thank J. Gonzalez-Espitia and K. Homicsko (Lausanne University 	<|special_separator|>
-(039.69, 325.55) (294.39, 325.55) (294.39, 332.57) (039.69, 332.57)      /T1_2 Hospital) for helpful discussions. This project received funding from the 	<|special_separator|>
-(039.69, 314.80) (295.67, 314.80) (295.67, 321.82) (039.69, 321.82)      /T1_2 EU project Fragment-Screen (Grant Agreement ID 101094131 to J.B.), the 	<|special_separator|>
-(039.69, 304.05) (284.69, 304.05) (284.69, 311.06) (039.69, 311.06)      /T1_2 European Union's Horizon 2020 research and innovation programme 	<|special_separator|>
-(039.69, 293.30) (290.65, 293.30) (290.65, 300.31) (039.69, 300.31)      /T1_2 (Marie Skłodowska-Curie Grant Agreement No. 955321 to A.D.) and the 	<|special_separator|>
-(039.69, 282.55) (262.68, 282.55) (262.68, 289.56) (039.69, 289.56)      /T1_2 University of Lausanne and Lausanne University Hospital (M.R.).	<|special_separator|>
-(039.69, 260.35) (149.49, 260.35) (149.49, 270.11) (039.69, 270.11)      /T1_5 Author contributions	<|special_separator|>
-(039.69, 250.31) (278.09, 250.31) (278.09, 257.32) (039.69, 257.32)      /T1_2 Conceptualization: B.H., M.R., J.B. Data curation: A.D., B.H. Formal 	<|special_separator|>
-(039.69, 239.55) (277.59, 239.55) (277.59, 246.57) (039.69, 246.57)      /T1_2 analysis: A.D., D.A.R., B.H. Funding acquisition: M.R. Investigation: 	<|special_separator|>
-(306.14, 734.06) (550.47, 734.06) (550.47, 741.07) (306.14, 741.07)      /T1_2 A.D., D.A.R., B.H. Methodology: B.H., J.B. Project administration: J.B. 	<|special_separator|>
-(306.14, 723.31) (560.79, 723.31) (560.79, 730.32) (306.14, 730.32)      /T1_2 Software: A.D., D.A.R., B.H., J.B. Supervision: M.R., J.B. Validation: M.R., 	<|special_separator|>
-(306.14, 712.55) (553.22, 712.55) (553.22, 719.57) (306.14, 719.57)      /T1_2 J.B. Visualization: A.D., D.A.R., B.H. Writing-original draft: A.D., M.R., 	<|special_separator|>
-(306.14, 701.80) (474.49, 701.80) (474.49, 708.82) (306.14, 708.82)      /T1_2 J.B. Writing-review and editing: A.D., M.R., J.B.	<|special_separator|>
-(306.14, 679.60) (410.67, 679.60) (410.67, 689.36) (306.14, 689.36)      /T1_5 Competing interests	<|special_separator|>
-(306.14, 669.56) (466.18, 669.56) (466.18, 676.57) (306.14, 676.57)      /T1_2 The authors declare no competing interests.	<|special_separator|>
-(306.14, 647.35) (425.20, 647.35) (425.20, 657.11) (306.14, 657.11)      /T1_5 Additional information	<|special_separator|>
-(306.14, 637.32) (360.06, 637.32) (360.06, 644.32) (306.14, 644.32)      /T1_4 Extended data	<|special_separator|>
-(359.92, 637.31) (463.62, 637.31) (463.62, 644.32) (359.92, 644.32)      /T1_2  is available for this paper at  	<|special_separator|>
-(306.14, 626.55) (468.32, 626.55) (468.32, 633.57) (306.14, 633.57)      /T1_2 https://doi.org/10.1038/s42256-026-01242-8.	<|special_separator|>
-(306.14, 605.07) (409.45, 605.07) (409.45, 612.06) (306.14, 612.06)      /T1_4 Supplementary information	<|special_separator|>
-(409.32, 605.05) (480.89, 605.05) (480.89, 612.06) (409.32, 612.06)      /T1_2  The online version  	<|special_separator|>
-(306.14, 594.30) (472.57, 594.30) (472.57, 601.31) (306.14, 601.31)      /T1_2 contains supplementary material available at  	<|special_separator|>
-(306.14, 583.55) (468.32, 583.55) (468.32, 590.56) (306.14, 590.56)      /T1_2 https://doi.org/10.1038/s42256-026-01242-8.	<|special_separator|>
-(306.14, 562.06) (468.97, 562.06) (468.97, 569.06) (306.14, 569.06)      /T1_4 Correspondence and requests for materials 	<|special_separator|>
-(468.51, 562.04) (555.74, 562.04) (555.74, 569.06) (468.51, 569.06)      /T1_2 should be addressed to 	<|special_separator|>
-(306.14, 551.29) (349.79, 551.29) (349.79, 558.30) (306.14, 558.30)      /T1_2 Jannis Born.	<|special_separator|>
-(306.14, 529.80) (394.71, 529.80) (394.71, 536.80) (306.14, 536.80)      /T1_4 Peer review information	<|special_separator|>
-(394.57, 529.79) (396.69, 529.79) (396.69, 536.80) (394.57, 536.80)      /T1_2  	<|special_separator|>
-(396.25, 529.79) (497.14, 529.79) (497.14, 536.80) (396.25, 536.80)      /T1_3 Nature Machine Intelligence	<|special_separator|>
-(497.03, 529.79) (548.05, 529.79) (548.05, 536.80) (497.03, 536.80)      /T1_2  thanks Di He, 	<|special_separator|>
-(306.14, 519.03) (546.22, 519.03) (546.22, 526.05) (306.14, 526.05)      /T1_2 Pingzhao Hu, Lei Li and the other, anonymous, reviewer(s) for their 	<|special_separator|>
-(306.14, 508.28) (463.67, 508.28) (463.67, 515.30) (306.14, 515.30)      /T1_2 contribution to the peer review of this work.	<|special_separator|>
-(306.14, 486.79) (446.61, 486.79) (446.61, 493.79) (306.14, 493.79)      /T1_4 Reprints and permissions information 	<|special_separator|>
-(446.15, 486.78) (498.45, 486.78) (498.45, 493.79) (446.15, 493.79)      /T1_2 is available at  	<|special_separator|>
-(306.13, 476.03) (401.37, 476.03) (401.37, 483.04) (306.13, 483.04)      /T1_2 www.nature.com/reprints.	<|special_separator|>
-(306.13, 454.54) (365.61, 454.54) (365.61, 461.54) (306.13, 461.54)      /T1_4 Publisher's note	<|special_separator|>
-(365.48, 454.52) (537.66, 454.52) (537.66, 461.54) (365.48, 461.54)      /T1_2  Springer Nature remains neutral with regard to 	<|special_separator|>
-(306.13, 443.77) (548.11, 443.77) (548.11, 450.78) (306.13, 450.78)      /T1_2 jurisdictional claims in published maps and institutional affiliations.	<|special_separator|>
-(306.13, 422.28) (356.37, 422.28) (356.37, 429.28) (306.13, 429.28)      /T1_4 Open Access	<|special_separator|>
-(356.36, 422.27) (544.50, 422.27) (544.50, 429.28) (356.36, 429.28)      /T1_2  This article is licensed under a Creative Commons 	<|special_separator|>
-(306.13, 411.51) (561.54, 411.51) (561.54, 418.53) (306.13, 418.53)      /T1_2 Attribution-NonCommercial-NoDerivatives 4.0 International License, 	<|special_separator|>
-(306.13, 400.76) (532.19, 400.76) (532.19, 407.78) (306.13, 407.78)      /T1_2 which permits any non-commercial use, sharing, distribution 	<|special_separator|>
-(306.13, 390.01) (538.46, 390.01) (538.46, 397.02) (306.13, 397.02)      /T1_2 and reproduction in any medium or format, as long as you give 	<|special_separator|>
-(306.13, 379.26) (558.83, 379.26) (558.83, 386.27) (306.13, 386.27)      /T1_2 appropriate credit to the original author(s) and the source, provide a 	<|special_separator|>
-(306.13, 368.51) (553.60, 368.51) (553.60, 375.52) (306.13, 375.52)      /T1_2 link to the Creative Commons licence, and indicate if you modified 	<|special_separator|>
-(306.13, 357.75) (559.53, 357.75) (559.53, 364.77) (306.13, 364.77)      /T1_2 the licensed material. You do not have permission under this licence 	<|special_separator|>
-(306.13, 347.00) (556.84, 347.00) (556.84, 354.02) (306.13, 354.02)      /T1_2 to share adapted material derived from this article or parts of it. The 	<|special_separator|>
-(306.13, 336.25) (558.73, 336.25) (558.73, 343.26) (306.13, 343.26)      /T1_2 images or other third party material in this article are included in the 	<|special_separator|>
-(306.13, 325.50) (556.61, 325.50) (556.61, 332.51) (306.13, 332.51)      /T1_2 article's Creative Commons licence, unless indicated otherwise in a 	<|special_separator|>
-(306.13, 314.75) (551.38, 314.75) (551.38, 321.76) (306.13, 321.76)      /T1_2 credit line to the material. If material is not included in the article's 	<|special_separator|>
-(306.13, 303.99) (553.31, 303.99) (553.31, 311.01) (306.13, 311.01)      /T1_2 Creative Commons licence and your intended use is not permitted 	<|special_separator|>
-(306.13, 293.24) (554.99, 293.24) (554.99, 300.26) (306.13, 300.26)      /T1_2 by statutory regulation or exceeds the permitted use, you will need 	<|special_separator|>
-(306.13, 282.49) (548.72, 282.49) (548.72, 289.50) (306.13, 289.50)      /T1_2 to obtain permission directly from the copyright holder. To view a 	<|special_separator|>
-(306.13, 271.74) (542.23, 271.74) (542.23, 278.75) (306.13, 278.75)      /T1_2 copy of this licence, visit http://creativecommons.org/licenses/ 	<|special_separator|>
-(306.13, 260.99) (359.08, 260.99) (359.08, 268.00) (306.13, 268.00)      /T1_2 by-nc-nd/4.0/.	<|special_separator|>
-(306.13, 239.48) (383.70, 239.48) (383.70, 246.50) (306.13, 246.50)      /T1_2 © The Author(s) 2026

groundtruth/parser/font_11.pdf.page_no_13.py.json.word.txt

@@ -1,539 +0,0 @@
-(039.69, 020.40) (268.42, 020.40) (268.42, 027.41) (039.69, 027.41)      /T1_0 NatureMachineIntelligence|Volume8|June2026|984-996	<|special_separator|>
-(546.10, 020.07) (562.05, 020.07) (562.05, 027.05) (546.10, 027.05)      /T1_1 996	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 758.00) (569.67, 758.00) (569.67, 764.99) (391.20, 764.99)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 734.06) (050.73, 734.06) (050.73, 741.07) (039.69, 741.07)      /T1_2 55.	<|special_separator|>
-(056.69, 734.06) (087.89, 734.06) (087.89, 741.07) (056.69, 741.07)      /T1_2 Mariella,	<|special_separator|>
-(089.49, 734.06) (097.43, 734.06) (097.43, 741.07) (089.49, 741.07)      /T1_2 N.	<|special_separator|>
-(099.03, 734.06) (106.31, 734.06) (106.31, 741.07) (099.03, 741.07)      /T1_2 et	<|special_separator|>
-(107.91, 734.06) (116.62, 734.06) (116.62, 741.07) (107.91, 741.07)      /T1_2 al.	<|special_separator|>
-(118.22, 734.06) (178.18, 734.06) (178.18, 741.07) (118.22, 741.07)      /T1_2 Quantumtheory	<|special_separator|>
-(179.78, 734.06) (193.69, 734.06) (193.69, 741.07) (179.78, 741.07)      /T1_2 and	<|special_separator|>
-(195.29, 734.06) (236.36, 734.06) (236.36, 741.07) (195.29, 741.07)      /T1_2 application	<|special_separator|>
-(237.96, 734.06) (245.46, 734.06) (245.46, 741.07) (237.96, 741.07)      /T1_2 of	<|special_separator|>
-(247.06, 734.06) (286.04, 734.06) (286.04, 741.07) (247.06, 741.07)      /T1_2 contextual	<|special_separator|>
-(056.69, 723.30) (084.74, 723.30) (084.74, 730.32) (056.69, 730.32)      /T1_2 optimal	<|special_separator|>
-(086.34, 723.30) (122.15, 723.30) (122.15, 730.32) (086.34, 730.32)      /T1_2 transport.	<|special_separator|>
-(123.75, 723.30) (130.65, 723.30) (130.65, 730.32) (123.75, 730.32)      /T1_2 In	<|special_separator|>
-(132.25, 723.30) (150.80, 723.30) (150.80, 730.32) (132.25, 730.32)      /T1_3 Proc.	<|special_separator|>
-(152.32, 723.30) (166.52, 723.30) (166.52, 730.32) (152.32, 730.32)      /T1_3 41st	<|special_separator|>
-(168.04, 723.30) (213.16, 723.30) (213.16, 730.32) (168.04, 730.32)      /T1_3 International	<|special_separator|>
-(214.68, 723.30) (267.69, 723.30) (267.69, 730.32) (214.68, 730.32)      /T1_3 Conferenceon	<|special_separator|>
-(056.69, 712.55) (087.35, 712.55) (087.35, 719.57) (056.69, 719.57)      /T1_3 Machine	<|special_separator|>
-(088.87, 712.55) (120.28, 712.55) (120.28, 719.57) (088.87, 719.57)      /T1_3 Learning	<|special_separator|>
-(120.14, 712.55) (122.39, 712.55) (122.39, 719.57) (120.14, 719.57)      /T1_2 ,	<|special_separator|>
-(123.99, 712.55) (137.87, 712.55) (137.87, 719.57) (123.99, 719.57)      /T1_2 Vol.	<|special_separator|>
-(139.47, 712.55) (155.06, 712.55) (155.06, 719.57) (139.47, 719.57)      /T1_2 235,	<|special_separator|>
-(156.66, 712.55) (207.07, 712.55) (207.07, 719.57) (156.66, 719.57)      /T1_2 34822-34845	<|special_separator|>
-(208.67, 712.55) (233.70, 712.55) (233.70, 719.57) (208.67, 719.57)      /T1_2 (PMLR,	<|special_separator|>
-(235.30, 712.55) (258.54, 712.55) (258.54, 719.57) (235.30, 719.57)      /T1_2 2024).	<|special_separator|>
-(039.69, 701.80) (051.11, 701.80) (051.11, 708.81) (039.69, 708.81)      /T1_2 56.	<|special_separator|>
-(056.69, 701.80) (076.42, 701.80) (076.42, 708.81) (056.69, 708.81)      /T1_2 Basu,	<|special_separator|>
-(078.02, 701.80) (084.95, 701.80) (084.95, 708.81) (078.02, 708.81)      /T1_2 S.	<|special_separator|>
-(086.55, 701.80) (093.83, 701.80) (093.83, 708.81) (086.55, 708.81)      /T1_2 et	<|special_separator|>
-(095.43, 701.80) (104.14, 701.80) (104.14, 708.81) (095.43, 708.81)      /T1_2 al.	<|special_separator|>
-(105.74, 701.80) (144.98, 701.80) (144.98, 708.81) (105.74, 708.81)      /T1_2 Advancing	<|special_separator|>
-(146.58, 701.80) (184.57, 701.80) (184.57, 708.81) (146.58, 708.81)      /T1_2 single-cell	<|special_separator|>
-(186.17, 701.80) (208.31, 701.80) (208.31, 708.81) (186.17, 708.81)      /T1_2 omics	<|special_separator|>
-(209.91, 701.80) (223.82, 701.80) (223.82, 708.81) (209.91, 708.81)      /T1_2 and	<|special_separator|>
-(225.42, 701.80) (263.69, 701.80) (263.69, 708.81) (225.42, 708.81)      /T1_2 cell-based	<|special_separator|>
-(056.69, 691.05) (102.60, 691.05) (102.60, 698.06) (056.69, 698.06)      /T1_2 therapeutics	<|special_separator|>
-(104.20, 691.05) (119.90, 691.05) (119.90, 698.06) (104.20, 698.06)      /T1_2 with	<|special_separator|>
-(121.50, 691.05) (154.62, 691.05) (154.62, 698.06) (121.50, 698.06)      /T1_2 quantum	<|special_separator|>
-(156.22, 691.05) (198.66, 691.05) (198.66, 698.06) (156.22, 698.06)      /T1_2 computing.	<|special_separator|>
-(200.26, 691.05) (214.85, 691.05) (214.85, 698.06) (200.26, 698.06)      /T1_3 Nat.	<|special_separator|>
-(216.37, 691.05) (231.12, 691.05) (231.12, 698.06) (216.37, 698.06)      /T1_3 Rev.	<|special_separator|>
-(232.64, 691.05) (248.41, 691.05) (248.41, 698.06) (232.64, 698.06)      /T1_3 Mol.	<|special_separator|>
-(249.93, 691.05) (264.38, 691.05) (264.38, 698.06) (249.93, 698.06)      /T1_3 Cell	<|special_separator|>
-(265.90, 691.05) (281.69, 691.05) (281.69, 698.06) (265.90, 698.06)      /T1_3 Biol.	<|special_separator|>
-(283.34, 691.06) (292.37, 691.06) (292.37, 698.06) (283.34, 698.06)      /T1_4 27	<|special_separator|>
-(292.24, 691.05) (294.49, 691.05) (294.49, 698.06) (292.24, 698.06)      /T1_2 ,	<|special_separator|>
-(056.69, 680.30) (089.90, 680.30) (089.90, 687.31) (056.69, 687.31)      /T1_2 394-408	<|special_separator|>
-(091.50, 680.30) (117.51, 680.30) (117.51, 687.31) (091.50, 687.31)      /T1_2 (2026).	<|special_separator|>
-(039.68, 669.54) (049.55, 669.54) (049.55, 676.56) (039.68, 676.56)      /T1_2 57.	<|special_separator|>
-(056.69, 669.54) (088.12, 669.54) (088.12, 676.56) (056.69, 676.56)      /T1_2 Ramdas,	<|special_separator|>
-(089.72, 669.54) (099.50, 669.54) (099.50, 676.56) (089.72, 676.56)      /T1_2 A.,	<|special_separator|>
-(101.10, 669.54) (125.21, 669.54) (125.21, 676.56) (101.10, 676.56)      /T1_2 Trillos,	<|special_separator|>
-(126.81, 669.54) (134.75, 669.54) (134.75, 676.56) (126.81, 676.56)      /T1_2 N.	<|special_separator|>
-(136.35, 669.54) (144.25, 669.54) (144.25, 676.56) (136.35, 676.56)      /T1_2 G.	<|special_separator|>
-(145.85, 669.54) (177.80, 669.54) (177.80, 676.56) (145.85, 676.56)      /T1_2 &Cuturi,	<|special_separator|>
-(179.40, 669.54) (188.40, 669.54) (188.40, 676.56) (179.40, 676.56)      /T1_2 M.	<|special_separator|>
-(190.00, 669.54) (246.48, 669.54) (246.48, 676.56) (190.00, 676.56)      /T1_2 OnWasserstein	<|special_separator|>
-(248.08, 669.54) (291.10, 669.54) (291.10, 676.56) (248.08, 676.56)      /T1_2 two-sample	<|special_separator|>
-(056.69, 658.79) (082.02, 658.79) (082.02, 665.81) (056.69, 665.81)      /T1_2 testing	<|special_separator|>
-(083.62, 658.79) (097.53, 658.79) (097.53, 665.81) (083.62, 665.81)      /T1_2 and	<|special_separator|>
-(099.13, 658.79) (125.00, 658.79) (125.00, 665.81) (099.13, 665.81)      /T1_2 related	<|special_separator|>
-(126.60, 658.79) (155.02, 658.79) (155.02, 665.81) (126.60, 665.81)      /T1_2 families	<|special_separator|>
-(156.62, 658.79) (164.11, 658.79) (164.11, 665.81) (156.62, 665.81)      /T1_2 of	<|special_separator|>
-(165.71, 658.79) (219.76, 658.79) (219.76, 665.81) (165.71, 665.81)      /T1_2 nonparametric	<|special_separator|>
-(221.36, 658.79) (241.05, 658.79) (241.05, 665.81) (221.36, 665.81)      /T1_2 tests.	<|special_separator|>
-(242.65, 658.79) (270.40, 658.79) (270.40, 665.81) (242.65, 665.81)      /T1_3 Entropy	<|special_separator|>
-(272.06, 658.81) (280.57, 658.81) (280.57, 665.81) (272.06, 665.81)      /T1_4 19	<|special_separator|>
-(280.44, 658.79) (282.69, 658.79) (282.69, 665.81) (280.44, 665.81)      /T1_2 ,	<|special_separator|>
-(056.69, 648.04) (065.57, 648.04) (065.57, 655.05) (056.69, 655.05)      /T1_2 47	<|special_separator|>
-(067.17, 648.04) (091.35, 648.04) (091.35, 655.05) (067.17, 655.05)      /T1_2 (2017).	<|special_separator|>
-(039.68, 637.29) (051.05, 637.29) (051.05, 644.30) (039.68, 644.30)      /T1_2 58.	<|special_separator|>
-(056.69, 637.29) (092.06, 637.29) (092.06, 644.30) (056.69, 644.30)      /T1_2 Salimans,	<|special_separator|>
-(093.66, 637.29) (101.50, 637.29) (101.50, 644.30) (093.66, 644.30)      /T1_2 T.,	<|special_separator|>
-(103.10, 637.29) (128.50, 637.29) (128.50, 644.30) (103.10, 644.30)      /T1_2 Zhang,	<|special_separator|>
-(130.10, 637.29) (140.24, 637.29) (140.24, 644.30) (130.10, 644.30)      /T1_2 H.,	<|special_separator|>
-(141.84, 637.29) (173.04, 637.29) (173.04, 644.30) (141.84, 644.30)      /T1_2 Radford,	<|special_separator|>
-(174.64, 637.29) (182.26, 637.29) (182.26, 644.30) (174.64, 644.30)      /T1_2 A.	<|special_separator|>
-(183.86, 637.29) (223.46, 637.29) (223.46, 644.30) (183.86, 644.30)      /T1_2 &Metaxas,	<|special_separator|>
-(225.06, 637.29) (232.30, 637.29) (232.30, 644.30) (225.06, 644.30)      /T1_2 D.	<|special_separator|>
-(233.90, 637.29) (293.43, 637.29) (293.43, 644.30) (233.90, 644.30)      /T1_2 ImprovingGANs	<|special_separator|>
-(056.69, 626.54) (076.71, 626.54) (076.71, 633.55) (056.69, 633.55)      /T1_2 using	<|special_separator|>
-(078.31, 626.54) (106.36, 626.54) (106.36, 633.55) (078.31, 633.55)      /T1_2 optimal	<|special_separator|>
-(107.96, 626.54) (143.77, 626.54) (143.77, 633.55) (107.96, 633.55)      /T1_2 transport.	<|special_separator|>
-(145.37, 626.54) (152.27, 626.54) (152.27, 633.55) (145.37, 633.55)      /T1_2 In	<|special_separator|>
-(153.87, 626.54) (198.98, 626.54) (198.98, 633.55) (153.87, 633.55)      /T1_3 International	<|special_separator|>
-(200.50, 626.54) (242.69, 626.54) (242.69, 633.55) (200.50, 633.55)      /T1_3 Conference	<|special_separator|>
-(244.21, 626.54) (286.44, 626.54) (286.44, 633.55) (244.21, 633.55)      /T1_3 onLearning	<|special_separator|>
-(056.69, 615.78) (114.53, 615.78) (114.53, 622.80) (056.69, 622.80)      /T1_3 Representations	<|special_separator|>
-(116.12, 615.78) (268.32, 615.78) (268.32, 622.80) (116.12, 622.80)      /T1_2 https://openreview.net/pdf?id=rkQkBnJAb	<|special_separator|>
-(269.92, 615.78) (294.60, 615.78) (294.60, 622.80) (269.92, 622.80)      /T1_2 (2018).	<|special_separator|>
-(039.68, 605.03) (050.87, 605.03) (050.87, 612.05) (039.68, 612.05)      /T1_2 59.	<|special_separator|>
-(056.69, 605.03) (097.81, 605.03) (097.81, 612.05) (056.69, 612.05)      /T1_2 Hendrycks,	<|special_separator|>
-(099.41, 605.03) (106.66, 605.03) (106.66, 612.05) (099.41, 612.05)      /T1_2 D.	<|special_separator|>
-(108.25, 605.03) (144.33, 605.03) (144.33, 612.05) (108.25, 612.05)      /T1_2 &Gimpel,	<|special_separator|>
-(145.93, 605.03) (153.01, 605.03) (153.01, 612.05) (145.93, 612.05)      /T1_2 K.	<|special_separator|>
-(154.61, 605.03) (187.92, 605.03) (187.92, 612.05) (154.61, 612.05)      /T1_2 Gaussian	<|special_separator|>
-(189.52, 605.03) (207.38, 605.03) (207.38, 612.05) (189.52, 612.05)      /T1_2 error	<|special_separator|>
-(208.98, 605.03) (229.55, 605.03) (229.55, 612.05) (208.98, 612.05)      /T1_2 linear	<|special_separator|>
-(231.15, 605.03) (248.99, 605.03) (248.99, 612.05) (231.15, 612.05)      /T1_2 units	<|special_separator|>
-(250.58, 605.03) (281.42, 605.03) (281.42, 612.05) (250.58, 612.05)      /T1_2 (GELUs).	<|special_separator|>
-(056.69, 594.28) (085.74, 594.28) (085.74, 601.29) (056.69, 601.29)      /T1_2 Preprint	<|special_separator|>
-(087.34, 594.28) (094.39, 594.28) (094.39, 601.29) (087.34, 601.29)      /T1_2 at	<|special_separator|>
-(095.99, 594.28) (210.26, 594.28) (210.26, 601.29) (095.99, 601.29)      /T1_2 http://arxiv.org/abs/1606.08415	<|special_separator|>
-(211.86, 594.28) (237.84, 594.28) (237.84, 601.29) (211.86, 601.29)      /T1_2 (2023).	<|special_separator|>
-(039.68, 583.53) (051.62, 583.53) (051.62, 590.54) (039.68, 590.54)      /T1_2 60.	<|special_separator|>
-(056.69, 583.53) (093.00, 583.53) (093.00, 590.54) (056.69, 590.54)      /T1_2 deLeeuw,	<|special_separator|>
-(094.52, 583.53) (099.76, 583.53) (099.76, 590.54) (094.52, 590.54)      /T1_2 J.	<|special_separator|>
-(101.28, 583.53) (143.02, 583.53) (143.02, 590.54) (101.28, 590.54)      /T1_2 Application	<|special_separator|>
-(144.54, 583.53) (152.00, 583.53) (152.00, 590.54) (144.54, 590.54)      /T1_2 of	<|special_separator|>
-(153.52, 583.53) (179.70, 583.53) (179.70, 590.54) (153.52, 590.54)      /T1_2 convex	<|special_separator|>
-(181.22, 583.53) (209.93, 583.53) (209.93, 590.54) (181.22, 590.54)      /T1_2 analysis	<|special_separator|>
-(211.45, 583.53) (218.98, 583.53) (218.98, 590.54) (211.45, 590.54)      /T1_2 to	<|special_separator|>
-(220.50, 583.53) (283.42, 583.53) (283.42, 590.54) (220.50, 590.54)      /T1_2 multidimensional	<|special_separator|>
-(056.69, 572.78) (084.84, 572.78) (084.84, 579.79) (056.69, 579.79)      /T1_2 scaling.	<|special_separator|>
-(086.38, 572.78) (219.02, 572.78) (219.02, 579.79) (086.38, 579.79)      /T1_3 UCLA:DepartmentofStatistics,UCLA	<|special_separator|>
-(220.55, 572.78) (294.51, 572.78) (294.51, 579.79) (220.55, 579.79)      /T1_2 https://escholarship.	<|special_separator|>
-(056.69, 562.02) (138.77, 562.02) (138.77, 569.04) (056.69, 569.04)      /T1_2 org/uc/item/4ps3b5mj	<|special_separator|>
-(140.29, 562.02) (163.14, 562.02) (163.14, 569.04) (140.29, 569.04)      /T1_2 (2011).	<|special_separator|>
-(039.68, 551.27) (049.72, 551.27) (049.72, 558.29) (039.68, 558.29)      /T1_2 61.	<|special_separator|>
-(056.69, 551.27) (083.81, 551.27) (083.81, 558.29) (056.69, 558.29)      /T1_2 Zaheer,	<|special_separator|>
-(085.41, 551.27) (094.41, 551.27) (094.41, 558.29) (085.41, 558.29)      /T1_2 M.	<|special_separator|>
-(096.01, 551.27) (103.29, 551.27) (103.29, 558.29) (096.01, 558.29)      /T1_2 et	<|special_separator|>
-(104.89, 551.27) (113.60, 551.27) (113.60, 558.29) (104.89, 558.29)      /T1_2 al.	<|special_separator|>
-(115.20, 551.27) (134.56, 551.27) (134.56, 558.29) (115.20, 558.29)      /T1_2 Deep	<|special_separator|>
-(136.16, 551.27) (153.06, 551.27) (153.06, 558.29) (136.16, 558.29)      /T1_2 sets.	<|special_separator|>
-(154.66, 551.27) (161.56, 551.27) (161.56, 558.29) (154.66, 558.29)      /T1_2 In	<|special_separator|>
-(163.16, 551.27) (177.30, 551.27) (177.30, 558.29) (163.16, 558.29)      /T1_3 31st	<|special_separator|>
-(178.82, 551.27) (231.83, 551.27) (231.83, 558.29) (178.82, 558.29)      /T1_3 Conferenceon	<|special_separator|>
-(056.69, 540.52) (080.44, 540.52) (080.44, 547.53) (056.69, 547.53)      /T1_3 Neural	<|special_separator|>
-(081.96, 540.52) (123.48, 540.52) (123.48, 547.53) (081.96, 547.53)      /T1_3 Information	<|special_separator|>
-(125.00, 540.52) (164.28, 540.52) (164.28, 547.53) (125.00, 547.53)      /T1_3 Processing	<|special_separator|>
-(165.80, 540.52) (195.91, 540.52) (195.91, 547.53) (165.80, 547.53)      /T1_3 Systems	<|special_separator|>
-(197.43, 540.52) (217.08, 540.52) (217.08, 547.53) (197.43, 547.53)      /T1_3 (NIPS	<|special_separator|>
-(218.60, 540.52) (237.92, 540.52) (237.92, 547.53) (218.60, 547.53)      /T1_3 2017)	<|special_separator|>
-(056.69, 529.77) (238.23, 529.77) (238.23, 536.78) (056.69, 536.78)      /T1_2 https://papers.nips.cc/paper_files/paper/2017/file/	<|special_separator|>
-(056.69, 519.02) (230.30, 519.02) (230.30, 526.03) (056.69, 526.03)      /T1_2 f22e4747da1aa27e363d86d40ff442fe-Paper.pdf	<|special_separator|>
-(231.90, 519.02) (256.08, 519.02) (256.08, 526.03) (231.90, 526.03)      /T1_2 (2017).	<|special_separator|>
-(039.68, 508.26) (051.04, 508.26) (051.04, 515.28) (039.68, 515.28)      /T1_2 62.	<|special_separator|>
-(056.69, 508.26) (075.11, 508.26) (075.11, 515.28) (056.69, 515.28)      /T1_2 Wolf,	<|special_separator|>
-(076.71, 508.26) (082.51, 508.26) (082.51, 515.28) (076.71, 515.28)      /T1_2 F.	<|special_separator|>
-(084.11, 508.26) (093.89, 508.26) (093.89, 515.28) (084.11, 515.28)      /T1_2 A.,	<|special_separator|>
-(095.49, 508.26) (126.72, 508.26) (126.72, 515.28) (095.49, 515.28)      /T1_2 Angerer,	<|special_separator|>
-(128.32, 508.26) (134.24, 508.26) (134.24, 515.28) (128.32, 515.28)      /T1_2 P.	<|special_separator|>
-(135.84, 508.26) (164.36, 508.26) (164.36, 515.28) (135.84, 515.28)      /T1_2 &Theis,	<|special_separator|>
-(165.96, 508.26) (171.76, 508.26) (171.76, 515.28) (165.96, 515.28)      /T1_2 F.	<|special_separator|>
-(173.36, 508.26) (178.64, 508.26) (178.64, 515.28) (173.36, 515.28)      /T1_2 J.	<|special_separator|>
-(180.24, 508.26) (212.52, 508.26) (212.52, 515.28) (180.24, 515.28)      /T1_2 SCANPY:	<|special_separator|>
-(214.12, 508.26) (254.47, 508.26) (254.47, 515.28) (214.12, 515.28)      /T1_2 large-scale	<|special_separator|>
-(256.07, 508.26) (294.06, 508.26) (294.06, 515.28) (256.07, 515.28)      /T1_2 single-cell	<|special_separator|>
-(056.69, 497.51) (075.27, 497.51) (075.27, 504.53) (056.69, 504.53)      /T1_2 gene	<|special_separator|>
-(076.87, 497.51) (116.41, 497.51) (116.41, 504.53) (076.87, 504.53)      /T1_2 expression	<|special_separator|>
-(118.01, 497.51) (134.24, 497.51) (134.24, 504.53) (118.01, 504.53)      /T1_2 data	<|special_separator|>
-(135.84, 497.51) (166.98, 497.51) (166.98, 504.53) (135.84, 504.53)      /T1_2 analysis.	<|special_separator|>
-(168.58, 497.51) (216.84, 497.51) (216.84, 504.53) (168.58, 504.53)      /T1_3 GenomeBiol.	<|special_separator|>
-(218.48, 497.53) (226.99, 497.53) (226.99, 504.53) (218.48, 504.53)      /T1_4 19	<|special_separator|>
-(226.85, 497.51) (229.11, 497.51) (229.11, 504.53) (226.85, 504.53)      /T1_2 ,	<|special_separator|>
-(230.71, 497.51) (238.49, 497.51) (238.49, 504.53) (230.71, 504.53)      /T1_2 15	<|special_separator|>
-(240.09, 497.51) (264.77, 497.51) (264.77, 504.53) (240.09, 504.53)      /T1_2 (2018).	<|special_separator|>
-(039.69, 486.76) (051.37, 486.76) (051.37, 493.77) (039.69, 493.77)      /T1_2 63.	<|special_separator|>
-(056.70, 486.76) (097.73, 486.76) (097.73, 493.77) (056.70, 493.77)      /T1_2 Loshchilov,	<|special_separator|>
-(099.33, 486.76) (103.70, 486.76) (103.70, 493.77) (099.33, 493.77)      /T1_2 I.	<|special_separator|>
-(105.30, 486.76) (137.37, 486.76) (137.37, 493.77) (105.30, 493.77)      /T1_2 &Hutter,	<|special_separator|>
-(138.97, 486.76) (144.77, 486.76) (144.77, 493.77) (138.97, 493.77)      /T1_2 F.	<|special_separator|>
-(146.37, 486.76) (186.71, 486.76) (186.71, 493.77) (146.37, 493.77)      /T1_2 Decoupled	<|special_separator|>
-(188.31, 486.76) (213.29, 486.76) (213.29, 493.77) (188.31, 493.77)      /T1_2 weight	<|special_separator|>
-(214.89, 486.76) (237.03, 486.76) (237.03, 493.77) (214.89, 493.77)      /T1_2 decay	<|special_separator|>
-(238.63, 486.76) (291.07, 486.76) (291.07, 493.77) (238.63, 493.77)      /T1_2 regularization.	<|special_separator|>
-(056.70, 476.01) (063.60, 476.01) (063.60, 483.02) (056.70, 483.02)      /T1_2 In	<|special_separator|>
-(065.20, 476.01) (110.31, 476.01) (110.31, 483.02) (065.20, 483.02)      /T1_3 International	<|special_separator|>
-(111.83, 476.01) (154.02, 476.01) (154.02, 483.02) (111.83, 483.02)      /T1_3 Conference	<|special_separator|>
-(155.54, 476.01) (197.78, 476.01) (197.78, 483.02) (155.54, 483.02)      /T1_3 onLearning	<|special_separator|>
-(199.30, 476.01) (257.13, 476.01) (257.13, 483.02) (199.30, 483.02)      /T1_3 Representations	<|special_separator|>
-(258.77, 476.01) (285.43, 476.01) (285.43, 483.02) (258.77, 483.02)      /T1_2 https://	<|special_separator|>
-(056.69, 465.26) (187.53, 465.26) (187.53, 472.27) (056.69, 472.27)      /T1_2 openreview.net/pdf?id=Bkg6RiCqY7	<|special_separator|>
-(189.13, 465.26) (213.91, 465.26) (213.91, 472.27) (189.13, 472.27)      /T1_2 (2019).	<|special_separator|>
-(039.68, 454.50) (051.07, 454.50) (051.07, 461.52) (039.68, 461.52)      /T1_2 64.	<|special_separator|>
-(056.69, 454.50) (283.22, 454.50) (283.22, 461.52) (056.69, 461.52)      /T1_2 Cuturi,M.etal.OptimalTransportTools(OTT):aJAXtoolboxforall	<|special_separator|>
-(056.69, 443.75) (078.48, 443.75) (078.48, 450.77) (056.69, 450.77)      /T1_2 things	<|special_separator|>
-(079.68, 443.75) (152.18, 443.75) (152.18, 450.77) (079.68, 450.77)      /T1_2 Wasserstein.Preprint	<|special_separator|>
-(153.38, 443.75) (160.24, 443.75) (160.24, 450.77) (153.38, 450.77)      /T1_2 at	<|special_separator|>
-(161.44, 443.75) (267.00, 443.75) (267.00, 450.77) (161.44, 450.77)      /T1_2 http://arxiv.org/abs/2201.12324	<|special_separator|>
-(268.20, 443.75) (292.57, 443.75) (292.57, 450.77) (268.20, 450.77)      /T1_2 (2022).	<|special_separator|>
-(039.68, 433.00) (051.06, 433.00) (051.06, 440.01) (039.68, 440.01)      /T1_2 65.	<|special_separator|>
-(056.69, 433.00) (086.69, 433.00) (086.69, 440.01) (056.69, 440.01)      /T1_2 Kingma,	<|special_separator|>
-(088.29, 433.00) (095.54, 433.00) (095.54, 440.01) (088.29, 440.01)      /T1_2 D.	<|special_separator|>
-(097.14, 433.00) (103.06, 433.00) (103.06, 440.01) (097.14, 440.01)      /T1_2 P.	<|special_separator|>
-(104.66, 433.00) (141.45, 433.00) (141.45, 440.01) (104.66, 440.01)      /T1_2 &Welling,	<|special_separator|>
-(143.05, 433.00) (152.05, 433.00) (152.05, 440.01) (143.05, 440.01)      /T1_2 M.	<|special_separator|>
-(153.65, 433.00) (208.35, 433.00) (208.35, 440.01) (153.65, 440.01)      /T1_2 Auto-encoding	<|special_separator|>
-(209.95, 433.00) (247.99, 433.00) (247.99, 440.01) (209.95, 440.01)      /T1_2 variational	<|special_separator|>
-(249.59, 433.00) (273.05, 433.00) (273.05, 440.01) (249.59, 440.01)      /T1_2 Bayes.	<|special_separator|>
-(056.69, 422.25) (085.74, 422.25) (085.74, 429.26) (056.69, 429.26)      /T1_2 Preprint	<|special_separator|>
-(087.34, 422.25) (094.39, 422.25) (094.39, 429.26) (087.34, 429.26)      /T1_2 at	<|special_separator|>
-(095.99, 422.25) (204.54, 422.25) (204.54, 429.26) (095.99, 429.26)      /T1_2 https://arxiv.org/abs/1312.6114	<|special_separator|>
-(206.14, 422.25) (230.88, 422.25) (230.88, 429.26) (206.14, 429.26)      /T1_2 (2013).	<|special_separator|>
-(039.68, 411.50) (051.32, 411.50) (051.32, 418.51) (039.68, 418.51)      /T1_2 66.	<|special_separator|>
-(056.69, 411.50) (082.23, 411.50) (082.23, 418.51) (056.69, 418.51)      /T1_2 Bunne,	<|special_separator|>
-(083.83, 411.50) (091.59, 411.50) (091.59, 418.51) (083.83, 418.51)      /T1_2 C.	<|special_separator|>
-(093.19, 411.50) (100.47, 411.50) (100.47, 418.51) (093.19, 418.51)      /T1_2 et	<|special_separator|>
-(102.07, 411.50) (110.78, 411.50) (110.78, 418.51) (102.07, 418.51)      /T1_2 al.	<|special_separator|>
-(112.38, 411.50) (143.80, 411.50) (143.80, 418.51) (112.38, 418.51)      /T1_2 Datasets	<|special_separator|>
-(145.40, 411.50) (152.90, 411.50) (152.90, 418.51) (145.40, 418.51)      /T1_2 of	<|special_separator|>
-(154.50, 411.50) (184.65, 411.50) (184.65, 418.51) (154.50, 418.51)      /T1_2 learning	<|special_separator|>
-(186.25, 411.50) (224.25, 411.50) (224.25, 418.51) (186.25, 418.51)      /T1_2 single-cell	<|special_separator|>
-(225.85, 411.50) (271.74, 411.50) (271.74, 418.51) (225.85, 418.51)      /T1_2 perturbation	<|special_separator|>
-(056.69, 400.74) (094.08, 400.74) (094.08, 407.76) (056.69, 407.76)      /T1_2 responses	<|special_separator|>
-(095.68, 400.74) (115.69, 400.74) (115.69, 407.76) (095.68, 407.76)      /T1_2 using	<|special_separator|>
-(117.29, 400.74) (140.49, 400.74) (140.49, 407.76) (117.29, 407.76)      /T1_2 neural	<|special_separator|>
-(142.09, 400.74) (170.14, 400.74) (170.14, 407.76) (142.09, 407.76)      /T1_2 optimal	<|special_separator|>
-(171.74, 400.74) (207.55, 400.74) (207.55, 407.76) (171.74, 407.76)      /T1_2 transport.	<|special_separator|>
-(209.22, 400.74) (248.19, 400.74) (248.19, 407.76) (209.22, 407.76)      /T1_3 ETHZürich	<|special_separator|>
-(249.76, 400.74) (289.67, 400.74) (289.67, 407.76) (249.76, 407.76)      /T1_2 https://doi.	<|special_separator|>
-(056.70, 389.99) (172.52, 389.99) (172.52, 397.01) (056.70, 397.01)      /T1_2 org/10.3929/ethz-b-000609681	<|special_separator|>
-(174.12, 389.99) (200.10, 389.99) (200.10, 397.01) (174.12, 397.01)      /T1_2 (2023).	<|special_separator|>
-(039.69, 379.24) (049.94, 379.24) (049.94, 386.25) (039.69, 386.25)      /T1_2 67.	<|special_separator|>
-(056.70, 379.24) (076.14, 379.24) (076.14, 386.25) (056.70, 386.25)      /T1_2 Born,	<|special_separator|>
-(077.74, 379.24) (085.17, 379.24) (085.17, 386.25) (077.74, 386.25)      /T1_2 J.,	<|special_separator|>
-(086.77, 379.24) (120.40, 379.24) (120.40, 386.25) (086.77, 386.25)      /T1_2 Driessen,	<|special_separator|>
-(122.00, 379.24) (129.63, 379.24) (129.63, 386.25) (122.00, 386.25)      /T1_2 A.	<|special_separator|>
-(131.23, 379.24) (171.97, 379.24) (171.97, 386.25) (131.23, 386.25)      /T1_2 &Harsanyi,	<|special_separator|>
-(173.57, 379.24) (180.70, 379.24) (180.70, 386.25) (173.57, 386.25)      /T1_2 B.	<|special_separator|>
-(182.30, 379.24) (285.05, 379.24) (285.05, 386.25) (182.30, 386.25)      /T1_2 AI4SCR/conditional-monge-	<|special_separator|>
-(056.70, 368.49) (072.84, 368.49) (072.84, 375.50) (056.70, 375.50)      /T1_2 gap.	<|special_separator|>
-(074.44, 368.49) (101.93, 368.49) (101.93, 375.50) (074.44, 375.50)      /T1_3 Zenodo	<|special_separator|>
-(103.50, 368.49) (251.95, 368.49) (251.95, 375.50) (103.50, 375.50)      /T1_2 https://doi.org/10.5281/zenodo.17854338	<|special_separator|>
-(253.55, 368.49) (279.38, 368.49) (279.38, 375.50) (253.55, 375.50)      /T1_2 (2025).	<|special_separator|>
-(039.69, 346.35) (140.85, 346.35) (140.85, 356.11) (039.69, 356.11)      /T1_5 Acknowledgements	<|special_separator|>
-(039.69, 336.31) (139.97, 336.31) (139.97, 343.32) (039.69, 343.32)      /T1_2 WethankJ.Gonzalez-Espitia	<|special_separator|>
-(141.41, 336.31) (275.88, 336.31) (275.88, 343.32) (141.41, 343.32)      /T1_2 andK.Homicsko(LausanneUniversity	<|special_separator|>
-(039.69, 325.55) (071.90, 325.55) (071.90, 332.57) (039.69, 332.57)      /T1_2 Hospital)	<|special_separator|>
-(073.34, 325.55) (083.39, 325.55) (083.39, 332.57) (073.34, 332.57)      /T1_2 for	<|special_separator|>
-(084.83, 325.55) (110.27, 325.55) (110.27, 332.57) (084.83, 332.57)      /T1_2 helpful	<|special_separator|>
-(111.71, 325.55) (155.33, 325.55) (155.33, 332.57) (111.71, 332.57)      /T1_2 discussions.	<|special_separator|>
-(156.77, 325.55) (171.37, 325.55) (171.37, 332.57) (156.77, 332.57)      /T1_2 This	<|special_separator|>
-(172.81, 325.55) (198.43, 325.55) (198.43, 332.57) (172.81, 332.57)      /T1_2 project	<|special_separator|>
-(199.87, 325.55) (230.87, 325.55) (230.87, 332.57) (199.87, 332.57)      /T1_2 received	<|special_separator|>
-(232.31, 325.55) (260.37, 325.55) (260.37, 332.57) (232.31, 332.57)      /T1_2 funding	<|special_separator|>
-(261.81, 325.55) (279.03, 325.55) (279.03, 332.57) (261.81, 332.57)      /T1_2 from	<|special_separator|>
-(280.47, 325.55) (292.27, 325.55) (292.27, 332.57) (280.47, 332.57)      /T1_2 the	<|special_separator|>
-(039.69, 314.80) (076.66, 314.80) (076.66, 321.82) (039.69, 321.82)      /T1_2 EUproject	<|special_separator|>
-(078.10, 314.80) (140.42, 314.80) (140.42, 321.82) (078.10, 321.82)      /T1_2 Fragment-Screen	<|special_separator|>
-(141.86, 314.80) (164.57, 314.80) (164.57, 321.82) (141.86, 321.82)      /T1_2 (Grant	<|special_separator|>
-(166.01, 314.80) (280.31, 314.80) (280.31, 321.82) (166.01, 321.82)      /T1_2 AgreementID101094131toJ.B.),	<|special_separator|>
-(281.75, 314.80) (293.55, 314.80) (293.55, 321.82) (281.75, 321.82)      /T1_2 the	<|special_separator|>
-(039.69, 304.05) (074.11, 304.05) (074.11, 311.06) (039.69, 311.06)      /T1_2 European	<|special_separator|>
-(075.55, 304.05) (282.57, 304.05) (282.57, 311.06) (075.55, 311.06)      /T1_2 Union'sHorizon2020researchandinnovationprogramme	<|special_separator|>
-(039.69, 293.30) (062.10, 293.30) (062.10, 300.31) (039.69, 300.31)      /T1_2 (Marie	<|special_separator|>
-(063.54, 293.30) (288.53, 293.30) (288.53, 300.31) (063.54, 300.31)      /T1_2 Skłodowska-CurieGrantAgreementNo.955321toA.D.)andthe	<|special_separator|>
-(039.69, 282.55) (075.01, 282.55) (075.01, 289.56) (039.69, 289.56)      /T1_2 University	<|special_separator|>
-(076.45, 282.55) (083.86, 282.55) (083.86, 289.56) (076.45, 289.56)      /T1_2 of	<|special_separator|>
-(085.30, 282.55) (119.64, 282.55) (119.64, 289.56) (085.30, 289.56)      /T1_2 Lausanne	<|special_separator|>
-(121.08, 282.55) (207.37, 282.55) (207.37, 289.56) (121.08, 289.56)      /T1_2 andLausanneUniversity	<|special_separator|>
-(208.81, 282.55) (238.63, 282.55) (238.63, 289.56) (208.81, 289.56)      /T1_2 Hospital	<|special_separator|>
-(240.07, 282.55) (262.68, 282.55) (262.68, 289.56) (240.07, 289.56)      /T1_2 (M.R.).	<|special_separator|>
-(039.69, 260.35) (149.49, 260.35) (149.49, 270.11) (039.69, 270.11)      /T1_5 Authorcontributions	<|special_separator|>
-(039.69, 250.31) (108.85, 250.31) (108.85, 257.32) (039.69, 257.32)      /T1_2 Conceptualization:	<|special_separator|>
-(110.45, 250.31) (127.62, 250.31) (127.62, 257.32) (110.45, 257.32)      /T1_2 B.H.,	<|special_separator|>
-(129.22, 250.31) (147.36, 250.31) (147.36, 257.32) (129.22, 257.32)      /T1_2 M.R.,	<|special_separator|>
-(148.96, 250.31) (161.26, 250.31) (161.26, 257.32) (148.96, 257.32)      /T1_2 J.B.	<|special_separator|>
-(162.86, 250.31) (179.64, 250.31) (179.64, 257.32) (162.86, 257.32)      /T1_2 Data	<|special_separator|>
-(181.24, 250.31) (213.84, 250.31) (213.84, 257.32) (181.24, 257.32)      /T1_2 curation:	<|special_separator|>
-(215.44, 250.31) (232.36, 250.31) (232.36, 257.32) (215.44, 257.32)      /T1_2 A.D.,	<|special_separator|>
-(233.96, 250.31) (248.98, 250.31) (248.98, 257.32) (233.96, 257.32)      /T1_2 B.H.	<|special_separator|>
-(250.58, 250.31) (275.97, 250.31) (275.97, 257.32) (250.58, 257.32)      /T1_2 Formal	<|special_separator|>
-(039.69, 239.55) (070.91, 239.55) (070.91, 246.57) (039.69, 246.57)      /T1_2 analysis:	<|special_separator|>
-(072.51, 239.55) (089.43, 239.55) (089.43, 246.57) (072.51, 246.57)      /T1_2 A.D.,	<|special_separator|>
-(091.03, 239.55) (115.02, 239.55) (115.02, 246.57) (091.03, 246.57)      /T1_2 D.A.R.,	<|special_separator|>
-(116.62, 239.55) (131.64, 239.55) (131.64, 246.57) (116.62, 246.57)      /T1_2 B.H.	<|special_separator|>
-(133.24, 239.55) (163.22, 239.55) (163.22, 246.57) (133.24, 246.57)      /T1_2 Funding	<|special_separator|>
-(164.82, 239.55) (207.07, 239.55) (207.07, 246.57) (164.82, 246.57)      /T1_2 acquisition:	<|special_separator|>
-(208.67, 239.55) (224.67, 239.55) (224.67, 246.57) (208.67, 246.57)      /T1_2 M.R.	<|special_separator|>
-(226.27, 239.55) (275.47, 239.55) (275.47, 246.57) (226.27, 246.57)      /T1_2 Investigation:	<|special_separator|>
-(306.14, 734.06) (323.06, 734.06) (323.06, 741.07) (306.14, 741.07)      /T1_2 A.D.,	<|special_separator|>
-(324.66, 734.06) (348.65, 734.06) (348.65, 741.07) (324.66, 741.07)      /T1_2 D.A.R.,	<|special_separator|>
-(350.25, 734.06) (365.27, 734.06) (365.27, 741.07) (350.25, 741.07)      /T1_2 B.H.	<|special_separator|>
-(366.87, 734.06) (418.02, 734.06) (418.02, 741.07) (366.87, 741.07)      /T1_2 Methodology:	<|special_separator|>
-(419.62, 734.06) (436.79, 734.06) (436.79, 741.07) (419.62, 741.07)      /T1_2 B.H.,	<|special_separator|>
-(438.39, 734.06) (450.69, 734.06) (450.69, 741.07) (438.39, 741.07)      /T1_2 J.B.	<|special_separator|>
-(452.29, 734.06) (478.02, 734.06) (478.02, 741.07) (452.29, 741.07)      /T1_2 Project	<|special_separator|>
-(479.62, 734.06) (534.45, 734.06) (534.45, 741.07) (479.62, 741.07)      /T1_2 administration:	<|special_separator|>
-(536.05, 734.06) (548.35, 734.06) (548.35, 741.07) (536.05, 741.07)      /T1_2 J.B.	<|special_separator|>
-(306.14, 723.31) (341.17, 723.31) (341.17, 730.32) (306.14, 730.32)      /T1_2 Software:	<|special_separator|>
-(342.77, 723.31) (359.69, 723.31) (359.69, 730.32) (342.77, 730.32)      /T1_2 A.D.,	<|special_separator|>
-(361.29, 723.31) (385.28, 723.31) (385.28, 730.32) (361.29, 730.32)      /T1_2 D.A.R.,	<|special_separator|>
-(386.88, 723.31) (404.05, 723.31) (404.05, 730.32) (386.88, 730.32)      /T1_2 B.H.,	<|special_separator|>
-(405.65, 723.31) (417.95, 723.31) (417.95, 730.32) (405.65, 730.32)      /T1_2 J.B.	<|special_separator|>
-(419.55, 723.31) (464.89, 723.31) (464.89, 730.32) (419.55, 730.32)      /T1_2 Supervision:	<|special_separator|>
-(466.49, 723.31) (484.64, 723.31) (484.64, 730.32) (466.49, 730.32)      /T1_2 M.R.,	<|special_separator|>
-(486.24, 723.31) (498.54, 723.31) (498.54, 730.32) (486.24, 730.32)      /T1_2 J.B.	<|special_separator|>
-(500.14, 723.31) (538.92, 723.31) (538.92, 730.32) (500.14, 730.32)      /T1_2 Validation:	<|special_separator|>
-(540.52, 723.31) (558.67, 723.31) (558.67, 730.32) (540.52, 730.32)      /T1_2 M.R.,	<|special_separator|>
-(306.14, 712.55) (318.44, 712.55) (318.44, 719.57) (306.14, 719.57)      /T1_2 J.B.	<|special_separator|>
-(320.04, 712.55) (368.24, 712.55) (368.24, 719.57) (320.04, 719.57)      /T1_2 Visualization:	<|special_separator|>
-(369.84, 712.55) (386.75, 712.55) (386.75, 719.57) (369.84, 719.57)      /T1_2 A.D.,	<|special_separator|>
-(388.35, 712.55) (412.35, 712.55) (412.35, 719.57) (388.35, 719.57)      /T1_2 D.A.R.,	<|special_separator|>
-(413.95, 712.55) (428.96, 712.55) (428.96, 719.57) (413.95, 719.57)      /T1_2 B.H.	<|special_separator|>
-(430.56, 712.55) (491.46, 712.55) (491.46, 719.57) (430.56, 719.57)      /T1_2 Writing-original	<|special_separator|>
-(493.06, 712.55) (512.84, 712.55) (512.84, 719.57) (493.06, 719.57)      /T1_2 draft:	<|special_separator|>
-(514.44, 712.55) (531.35, 712.55) (531.35, 719.57) (514.44, 719.57)      /T1_2 A.D.,	<|special_separator|>
-(532.95, 712.55) (551.10, 712.55) (551.10, 719.57) (532.95, 719.57)      /T1_2 M.R.,	<|special_separator|>
-(306.14, 701.80) (318.44, 701.80) (318.44, 708.82) (306.14, 708.82)      /T1_2 J.B.	<|special_separator|>
-(320.04, 701.80) (377.33, 701.80) (377.33, 708.82) (320.04, 708.82)      /T1_2 Writing-review	<|special_separator|>
-(378.93, 701.80) (392.84, 701.80) (392.84, 708.82) (378.93, 708.82)      /T1_2 and	<|special_separator|>
-(394.44, 701.80) (422.33, 701.80) (422.33, 708.82) (394.44, 708.82)      /T1_2 editing:	<|special_separator|>
-(423.93, 701.80) (440.85, 701.80) (440.85, 708.82) (423.93, 708.82)      /T1_2 A.D.,	<|special_separator|>
-(442.45, 701.80) (460.60, 701.80) (460.60, 708.82) (442.45, 708.82)      /T1_2 M.R.,	<|special_separator|>
-(462.20, 701.80) (474.49, 701.80) (474.49, 708.82) (462.20, 708.82)      /T1_2 J.B.	<|special_separator|>
-(306.14, 679.60) (410.67, 679.60) (410.67, 689.36) (306.14, 689.36)      /T1_5 Competinginterests	<|special_separator|>
-(306.14, 669.56) (319.75, 669.56) (319.75, 676.57) (306.14, 676.57)      /T1_2 The	<|special_separator|>
-(321.35, 669.56) (349.06, 669.56) (349.06, 676.57) (321.35, 676.57)      /T1_2 authors	<|special_separator|>
-(350.66, 669.56) (378.26, 669.56) (378.26, 676.57) (350.66, 676.57)      /T1_2 declare	<|special_separator|>
-(379.86, 669.56) (389.37, 669.56) (389.37, 676.57) (379.86, 676.57)      /T1_2 no	<|special_separator|>
-(390.97, 669.56) (431.05, 669.56) (431.05, 676.57) (390.97, 676.57)      /T1_2 competing	<|special_separator|>
-(432.65, 669.56) (466.18, 669.56) (466.18, 676.57) (432.65, 676.57)      /T1_2 interests.	<|special_separator|>
-(306.14, 647.35) (425.20, 647.35) (425.20, 657.11) (306.14, 657.11)      /T1_5 Additionalinformation	<|special_separator|>
-(306.14, 637.32) (360.06, 637.32) (360.06, 644.32) (306.14, 644.32)      /T1_4 Extendeddata	<|special_separator|>
-(361.63, 637.31) (367.40, 637.31) (367.40, 644.32) (361.63, 644.32)      /T1_2 is	<|special_separator|>
-(369.00, 637.31) (401.54, 637.31) (401.54, 644.32) (369.00, 644.32)      /T1_2 available	<|special_separator|>
-(403.14, 637.31) (413.34, 637.31) (413.34, 644.32) (403.14, 644.32)      /T1_2 for	<|special_separator|>
-(414.94, 637.31) (428.13, 637.31) (428.13, 644.32) (414.94, 644.32)      /T1_2 this	<|special_separator|>
-(429.73, 637.31) (451.22, 637.31) (451.22, 644.32) (429.73, 644.32)      /T1_2 paper	<|special_separator|>
-(452.82, 637.31) (459.88, 637.31) (459.88, 644.32) (452.82, 644.32)      /T1_2 at	<|special_separator|>
-(306.14, 626.55) (468.32, 626.55) (468.32, 633.57) (306.14, 633.57)      /T1_2 https://doi.org/10.1038/s42256-026-01242-8.	<|special_separator|>
-(306.14, 605.07) (409.45, 605.07) (409.45, 612.06) (306.14, 612.06)      /T1_4 Supplementaryinformation	<|special_separator|>
-(411.02, 605.05) (424.63, 605.05) (424.63, 612.06) (411.02, 612.06)      /T1_2 The	<|special_separator|>
-(426.23, 605.05) (449.04, 605.05) (449.04, 612.06) (426.23, 612.06)      /T1_2 online	<|special_separator|>
-(450.64, 605.05) (477.18, 605.05) (477.18, 612.06) (450.64, 612.06)      /T1_2 version	<|special_separator|>
-(306.14, 594.30) (337.46, 594.30) (337.46, 601.31) (306.14, 601.31)      /T1_2 contains	<|special_separator|>
-(339.06, 594.30) (394.41, 594.30) (394.41, 601.31) (339.06, 601.31)      /T1_2 supplementary	<|special_separator|>
-(396.01, 594.30) (425.98, 594.30) (425.98, 601.31) (396.01, 601.31)      /T1_2 material	<|special_separator|>
-(427.58, 594.30) (460.12, 594.30) (460.12, 601.31) (427.58, 601.31)      /T1_2 available	<|special_separator|>
-(461.72, 594.30) (468.77, 594.30) (468.77, 601.31) (461.72, 601.31)      /T1_2 at	<|special_separator|>
-(306.14, 583.55) (468.32, 583.55) (468.32, 590.56) (306.14, 590.56)      /T1_2 https://doi.org/10.1038/s42256-026-01242-8.	<|special_separator|>
-(306.14, 562.06) (467.33, 562.06) (467.33, 569.06) (306.14, 569.06)      /T1_4 Correspondenceandrequestsformaterials	<|special_separator|>
-(468.51, 562.04) (493.56, 562.04) (493.56, 569.06) (468.51, 569.06)      /T1_2 should	<|special_separator|>
-(495.16, 562.04) (504.61, 562.04) (504.61, 569.06) (495.16, 569.06)      /T1_2 be	<|special_separator|>
-(506.21, 562.04) (544.45, 562.04) (544.45, 569.06) (506.21, 569.06)      /T1_2 addressed	<|special_separator|>
-(546.05, 562.04) (553.62, 562.04) (553.62, 569.06) (546.05, 569.06)      /T1_2 to	<|special_separator|>
-(306.14, 551.29) (328.75, 551.29) (328.75, 558.30) (306.14, 558.30)      /T1_2 Jannis	<|special_separator|>
-(330.35, 551.29) (349.79, 551.29) (349.79, 558.30) (330.35, 558.30)      /T1_2 Born.	<|special_separator|>
-(306.14, 529.80) (394.71, 529.80) (394.71, 536.80) (306.14, 536.80)      /T1_4 Peerreviewinformation	<|special_separator|>
-(396.25, 529.79) (420.57, 529.79) (420.57, 536.80) (396.25, 536.80)      /T1_3 Nature	<|special_separator|>
-(422.17, 529.79) (453.07, 529.79) (453.07, 536.80) (422.17, 536.80)      /T1_3 Machine	<|special_separator|>
-(454.67, 529.79) (497.14, 529.79) (497.14, 536.80) (454.67, 536.80)      /T1_3 Intelligence	<|special_separator|>
-(498.74, 529.79) (522.83, 529.79) (522.83, 536.80) (498.74, 536.80)      /T1_2 thanks	<|special_separator|>
-(524.43, 529.79) (531.94, 529.79) (531.94, 536.80) (524.43, 536.80)      /T1_2 Di	<|special_separator|>
-(533.54, 529.79) (545.93, 529.79) (545.93, 536.80) (533.54, 536.80)      /T1_2 He,	<|special_separator|>
-(306.14, 519.03) (339.74, 519.03) (339.74, 526.05) (306.14, 526.05)      /T1_2 Pingzhao	<|special_separator|>
-(341.34, 519.03) (353.98, 519.03) (353.98, 526.05) (341.34, 526.05)      /T1_2 Hu,	<|special_separator|>
-(355.58, 519.03) (366.16, 519.03) (366.16, 526.05) (355.58, 526.05)      /T1_2 Lei	<|special_separator|>
-(367.76, 519.03) (373.98, 519.03) (373.98, 526.05) (367.76, 526.05)      /T1_2 Li	<|special_separator|>
-(375.58, 519.03) (389.49, 519.03) (389.49, 526.05) (375.58, 526.05)      /T1_2 and	<|special_separator|>
-(391.09, 519.03) (403.05, 519.03) (403.05, 526.05) (391.09, 526.05)      /T1_2 the	<|special_separator|>
-(404.65, 519.03) (425.92, 519.03) (425.92, 526.05) (404.65, 526.05)      /T1_2 other,	<|special_separator|>
-(427.52, 519.03) (472.38, 519.03) (472.38, 526.05) (427.52, 526.05)      /T1_2 anonymous,	<|special_separator|>
-(473.98, 519.03) (513.90, 519.03) (513.90, 526.05) (473.98, 526.05)      /T1_2 reviewer(s)	<|special_separator|>
-(515.50, 519.03) (525.70, 519.03) (525.70, 526.05) (515.50, 526.05)      /T1_2 for	<|special_separator|>
-(527.30, 519.03) (544.10, 519.03) (544.10, 526.05) (527.30, 526.05)      /T1_2 their	<|special_separator|>
-(306.14, 508.28) (351.33, 508.28) (351.33, 515.30) (306.14, 515.30)      /T1_2 contribution	<|special_separator|>
-(352.93, 508.28) (360.49, 508.28) (360.49, 515.30) (352.93, 515.30)      /T1_2 to	<|special_separator|>
-(362.09, 508.28) (374.05, 508.28) (374.05, 515.30) (362.09, 515.30)      /T1_2 the	<|special_separator|>
-(375.65, 508.28) (392.45, 508.28) (392.45, 515.30) (375.65, 515.30)      /T1_2 peer	<|special_separator|>
-(394.05, 508.28) (418.10, 508.28) (418.10, 515.30) (394.05, 515.30)      /T1_2 review	<|special_separator|>
-(419.70, 508.28) (427.20, 508.28) (427.20, 515.30) (419.70, 515.30)      /T1_2 of	<|special_separator|>
-(428.80, 508.28) (441.99, 508.28) (441.99, 515.30) (428.80, 515.30)      /T1_2 this	<|special_separator|>
-(443.59, 508.28) (463.67, 508.28) (463.67, 515.30) (443.59, 515.30)      /T1_2 work.	<|special_separator|>
-(306.14, 486.79) (444.97, 486.79) (444.97, 493.79) (306.14, 493.79)      /T1_4 Reprintsandpermissionsinformation	<|special_separator|>
-(446.15, 486.78) (451.93, 486.78) (451.93, 493.79) (446.15, 493.79)      /T1_2 is	<|special_separator|>
-(453.53, 486.78) (486.06, 486.78) (486.06, 493.79) (453.53, 493.79)      /T1_2 available	<|special_separator|>
-(487.66, 486.78) (494.72, 486.78) (494.72, 493.79) (487.66, 493.79)      /T1_2 at	<|special_separator|>
-(306.13, 476.03) (401.37, 476.03) (401.37, 483.04) (306.13, 483.04)      /T1_2 www.nature.com/reprints.	<|special_separator|>
-(306.13, 454.54) (365.61, 454.54) (365.61, 461.54) (306.13, 461.54)      /T1_4 Publisher'snote	<|special_separator|>
-(367.18, 454.52) (398.89, 454.52) (398.89, 461.54) (367.18, 461.54)      /T1_2 Springer	<|special_separator|>
-(400.49, 454.52) (425.10, 454.52) (425.10, 461.54) (400.49, 461.54)      /T1_2 Nature	<|special_separator|>
-(426.70, 454.52) (455.78, 454.52) (455.78, 461.54) (426.70, 461.54)      /T1_2 remains	<|special_separator|>
-(457.38, 454.52) (483.31, 454.52) (483.31, 461.54) (457.38, 461.54)      /T1_2 neutral	<|special_separator|>
-(484.91, 454.52) (500.61, 454.52) (500.61, 461.54) (484.91, 461.54)      /T1_2 with	<|special_separator|>
-(502.21, 454.52) (526.37, 454.52) (526.37, 461.54) (502.21, 461.54)      /T1_2 regard	<|special_separator|>
-(527.97, 454.52) (535.54, 454.52) (535.54, 461.54) (527.97, 461.54)      /T1_2 to	<|special_separator|>
-(306.13, 443.77) (353.24, 443.77) (353.24, 450.78) (306.13, 450.78)      /T1_2 jurisdictional	<|special_separator|>
-(354.84, 443.77) (378.70, 443.77) (378.70, 450.78) (354.84, 450.78)      /T1_2 claims	<|special_separator|>
-(380.30, 443.77) (387.02, 443.77) (387.02, 450.78) (380.30, 450.78)      /T1_2 in	<|special_separator|>
-(388.62, 443.77) (425.14, 443.77) (425.14, 450.78) (388.62, 450.78)      /T1_2 published	<|special_separator|>
-(426.74, 443.77) (507.05, 443.77) (507.05, 450.78) (426.74, 450.78)      /T1_2 mapsandinstitutional	<|special_separator|>
-(508.65, 443.77) (548.11, 443.77) (548.11, 450.78) (508.65, 450.78)      /T1_2 affiliations.	<|special_separator|>
-(306.13, 422.28) (356.37, 422.28) (356.37, 429.28) (306.13, 429.28)      /T1_4 OpenAccess	<|special_separator|>
-(358.14, 422.27) (373.22, 422.27) (373.22, 429.28) (358.14, 429.28)      /T1_2 This	<|special_separator|>
-(374.98, 422.27) (398.54, 422.27) (398.54, 429.28) (374.98, 429.28)      /T1_2 article	<|special_separator|>
-(400.30, 422.27) (406.15, 422.27) (406.15, 429.28) (400.30, 429.28)      /T1_2 is	<|special_separator|>
-(407.91, 422.27) (439.42, 422.27) (439.42, 429.28) (407.91, 429.28)      /T1_2 licensed	<|special_separator|>
-(441.18, 422.27) (463.18, 422.27) (463.18, 429.28) (441.18, 429.28)      /T1_2 under	<|special_separator|>
-(464.94, 422.27) (469.26, 422.27) (469.26, 429.28) (464.94, 429.28)      /T1_2 a	<|special_separator|>
-(471.02, 422.27) (542.38, 422.27) (542.38, 429.28) (471.02, 429.28)      /T1_2 CreativeCommons	<|special_separator|>
-(306.13, 411.51) (464.01, 411.51) (464.01, 418.53) (306.13, 418.53)      /T1_2 Attribution-NonCommercial-NoDerivatives	<|special_separator|>
-(465.77, 411.51) (477.89, 411.51) (477.89, 418.53) (465.77, 418.53)      /T1_2 4.0	<|special_separator|>
-(479.65, 411.51) (527.02, 411.51) (527.02, 418.53) (479.65, 418.53)      /T1_2 International	<|special_separator|>
-(528.78, 411.51) (559.42, 411.51) (559.42, 418.53) (528.78, 418.53)      /T1_2 License,	<|special_separator|>
-(306.13, 400.76) (328.56, 400.76) (328.56, 407.78) (306.13, 407.78)      /T1_2 which	<|special_separator|>
-(330.32, 400.76) (358.74, 400.76) (358.74, 407.78) (330.32, 407.78)      /T1_2 permits	<|special_separator|>
-(360.50, 400.76) (373.67, 400.76) (373.67, 407.78) (360.50, 407.78)      /T1_2 any	<|special_separator|>
-(375.43, 400.76) (436.84, 400.76) (436.84, 407.78) (375.43, 407.78)      /T1_2 non-commercial	<|special_separator|>
-(438.60, 400.76) (453.89, 400.76) (453.89, 407.78) (438.60, 407.78)      /T1_2 use,	<|special_separator|>
-(455.65, 400.76) (485.53, 400.76) (485.53, 407.78) (455.65, 407.78)      /T1_2 sharing,	<|special_separator|>
-(487.29, 400.76) (530.07, 400.76) (530.07, 407.78) (487.29, 407.78)      /T1_2 distribution	<|special_separator|>
-(306.13, 390.01) (320.21, 390.01) (320.21, 397.02) (306.13, 397.02)      /T1_2 and	<|special_separator|>
-(321.97, 390.01) (370.70, 390.01) (370.70, 397.02) (321.97, 397.02)      /T1_2 reproduction	<|special_separator|>
-(372.46, 390.01) (379.26, 390.01) (379.26, 397.02) (372.46, 397.02)      /T1_2 in	<|special_separator|>
-(381.02, 390.01) (394.19, 390.01) (394.19, 397.02) (381.02, 397.02)      /T1_2 any	<|special_separator|>
-(395.95, 390.01) (426.74, 390.01) (426.74, 397.02) (395.95, 397.02)      /T1_2 medium	<|special_separator|>
-(428.50, 390.01) (436.31, 390.01) (436.31, 397.02) (428.50, 397.02)      /T1_2 or	<|special_separator|>
-(438.07, 390.01) (465.04, 390.01) (465.04, 397.02) (438.07, 397.02)      /T1_2 format,	<|special_separator|>
-(466.79, 390.01) (474.93, 390.01) (474.93, 397.02) (466.79, 397.02)      /T1_2 as	<|special_separator|>
-(476.69, 390.01) (493.59, 390.01) (493.59, 397.02) (476.69, 397.02)      /T1_2 long	<|special_separator|>
-(495.35, 390.01) (503.49, 390.01) (503.49, 397.02) (495.35, 397.02)      /T1_2 as	<|special_separator|>
-(505.25, 390.01) (518.90, 390.01) (518.90, 397.02) (505.25, 397.02)      /T1_2 you	<|special_separator|>
-(520.66, 390.01) (536.34, 390.01) (536.34, 397.02) (520.66, 397.02)      /T1_2 give	<|special_separator|>
-(306.13, 379.26) (349.74, 379.26) (349.74, 386.27) (306.13, 386.27)      /T1_2 appropriate	<|special_separator|>
-(351.50, 379.26) (373.27, 379.26) (373.27, 386.27) (351.50, 386.27)      /T1_2 credit	<|special_separator|>
-(375.03, 379.26) (382.68, 379.26) (382.68, 386.27) (375.03, 386.27)      /T1_2 to	<|special_separator|>
-(384.44, 379.26) (396.56, 379.26) (396.56, 386.27) (384.44, 386.27)      /T1_2 the	<|special_separator|>
-(398.32, 379.26) (426.53, 379.26) (426.53, 386.27) (398.32, 386.27)      /T1_2 original	<|special_separator|>
-(428.29, 379.26) (461.55, 379.26) (461.55, 386.27) (428.29, 386.27)      /T1_2 author(s)	<|special_separator|>
-(463.31, 379.26) (477.38, 379.26) (477.38, 386.27) (463.31, 386.27)      /T1_2 and	<|special_separator|>
-(479.14, 379.26) (491.26, 379.26) (491.26, 386.27) (479.14, 386.27)      /T1_2 the	<|special_separator|>
-(493.02, 379.26) (520.52, 379.26) (520.52, 386.27) (493.02, 386.27)      /T1_2 source,	<|special_separator|>
-(522.28, 379.26) (550.63, 379.26) (550.63, 386.27) (522.28, 386.27)      /T1_2 provide	<|special_separator|>
-(552.39, 379.26) (556.71, 379.26) (556.71, 386.27) (552.39, 386.27)      /T1_2 a	<|special_separator|>
-(306.13, 368.51) (319.36, 368.51) (319.36, 375.52) (306.13, 375.52)      /T1_2 link	<|special_separator|>
-(321.12, 368.51) (328.77, 368.51) (328.77, 375.52) (321.12, 375.52)      /T1_2 to	<|special_separator|>
-(330.53, 368.51) (342.65, 368.51) (342.65, 375.52) (330.53, 375.52)      /T1_2 the	<|special_separator|>
-(344.41, 368.51) (375.45, 368.51) (375.45, 375.52) (344.41, 375.52)      /T1_2 Creative	<|special_separator|>
-(377.21, 368.51) (446.94, 368.51) (446.94, 375.52) (377.21, 375.52)      /T1_2 Commonslicence,	<|special_separator|>
-(448.70, 368.51) (462.78, 368.51) (462.78, 375.52) (448.70, 375.52)      /T1_2 and	<|special_separator|>
-(464.54, 368.51) (494.51, 368.51) (494.51, 375.52) (464.54, 375.52)      /T1_2 indicate	<|special_separator|>
-(496.27, 368.51) (501.06, 368.51) (501.06, 375.52) (496.27, 375.52)      /T1_2 if	<|special_separator|>
-(502.82, 368.51) (516.47, 368.51) (516.47, 375.52) (502.82, 375.52)      /T1_2 you	<|special_separator|>
-(518.23, 368.51) (551.48, 368.51) (551.48, 375.52) (518.23, 375.52)      /T1_2 modified	<|special_separator|>
-(306.13, 357.75) (318.25, 357.75) (318.25, 364.77) (306.13, 364.77)      /T1_2 the	<|special_separator|>
-(320.01, 357.75) (351.52, 357.75) (351.52, 364.77) (320.01, 364.77)      /T1_2 licensed	<|special_separator|>
-(353.28, 357.75) (386.04, 357.75) (386.04, 364.77) (353.28, 364.77)      /T1_2 material.	<|special_separator|>
-(387.80, 357.75) (401.56, 357.75) (401.56, 364.77) (387.80, 364.77)      /T1_2 You	<|special_separator|>
-(403.32, 357.75) (413.14, 357.75) (413.14, 364.77) (403.32, 364.77)      /T1_2 do	<|special_separator|>
-(414.90, 357.75) (427.31, 357.75) (427.31, 364.77) (414.90, 364.77)      /T1_2 not	<|special_separator|>
-(429.07, 357.75) (446.69, 357.75) (446.69, 364.77) (429.07, 364.77)      /T1_2 have	<|special_separator|>
-(448.45, 357.75) (489.45, 357.75) (489.45, 364.77) (448.45, 364.77)      /T1_2 permission	<|special_separator|>
-(491.21, 357.75) (513.22, 357.75) (513.22, 364.77) (491.21, 364.77)      /T1_2 under	<|special_separator|>
-(514.98, 357.75) (528.41, 357.75) (528.41, 364.77) (514.98, 364.77)      /T1_2 this	<|special_separator|>
-(530.17, 357.75) (557.41, 357.75) (557.41, 364.77) (530.17, 364.77)      /T1_2 licence	<|special_separator|>
-(306.13, 347.00) (313.78, 347.00) (313.78, 354.02) (306.13, 354.02)      /T1_2 to	<|special_separator|>
-(315.54, 347.00) (335.82, 347.00) (335.82, 354.02) (315.54, 354.02)      /T1_2 share	<|special_separator|>
-(337.58, 347.00) (368.51, 347.00) (368.51, 354.02) (337.58, 354.02)      /T1_2 adapted	<|special_separator|>
-(370.27, 347.00) (400.80, 347.00) (400.80, 354.02) (370.27, 354.02)      /T1_2 material	<|special_separator|>
-(402.56, 347.00) (430.74, 347.00) (430.74, 354.02) (402.56, 354.02)      /T1_2 derived	<|special_separator|>
-(432.50, 347.00) (450.20, 347.00) (450.20, 354.02) (432.50, 354.02)      /T1_2 from	<|special_separator|>
-(451.96, 347.00) (465.39, 347.00) (465.39, 354.02) (451.96, 354.02)      /T1_2 this	<|special_separator|>
-(467.15, 347.00) (490.71, 347.00) (490.71, 354.02) (467.15, 354.02)      /T1_2 article	<|special_separator|>
-(492.47, 347.00) (500.28, 347.00) (500.28, 354.02) (492.47, 354.02)      /T1_2 or	<|special_separator|>
-(502.04, 347.00) (521.01, 347.00) (521.01, 354.02) (502.04, 354.02)      /T1_2 parts	<|special_separator|>
-(522.77, 347.00) (530.35, 347.00) (530.35, 354.02) (522.77, 354.02)      /T1_2 of	<|special_separator|>
-(532.11, 347.00) (539.20, 347.00) (539.20, 354.02) (532.11, 354.02)      /T1_2 it.	<|special_separator|>
-(540.96, 347.00) (554.72, 347.00) (554.72, 354.02) (540.96, 354.02)      /T1_2 The	<|special_separator|>
-(306.13, 336.25) (333.10, 336.25) (333.10, 343.26) (306.13, 343.26)      /T1_2 images	<|special_separator|>
-(334.86, 336.25) (342.67, 336.25) (342.67, 343.26) (334.86, 343.26)      /T1_2 or	<|special_separator|>
-(344.43, 336.25) (364.34, 336.25) (364.34, 343.26) (344.43, 343.26)      /T1_2 other	<|special_separator|>
-(366.10, 336.25) (383.58, 336.25) (383.58, 343.26) (366.10, 343.26)      /T1_2 third	<|special_separator|>
-(385.34, 336.25) (404.70, 336.25) (404.70, 343.26) (385.34, 343.26)      /T1_2 party	<|special_separator|>
-(406.46, 336.25) (436.99, 336.25) (436.99, 343.26) (406.46, 343.26)      /T1_2 material	<|special_separator|>
-(438.75, 336.25) (445.55, 336.25) (445.55, 343.26) (438.75, 343.26)      /T1_2 in	<|special_separator|>
-(447.31, 336.25) (460.74, 336.25) (460.74, 343.26) (447.31, 343.26)      /T1_2 this	<|special_separator|>
-(462.50, 336.25) (486.06, 336.25) (486.06, 343.26) (462.50, 343.26)      /T1_2 article	<|special_separator|>
-(487.82, 336.25) (499.52, 336.25) (499.52, 343.26) (487.82, 343.26)      /T1_2 are	<|special_separator|>
-(501.28, 336.25) (534.17, 336.25) (534.17, 343.26) (501.28, 343.26)      /T1_2 included	<|special_separator|>
-(535.93, 336.25) (542.73, 336.25) (542.73, 343.26) (535.93, 343.26)      /T1_2 in	<|special_separator|>
-(544.49, 336.25) (556.61, 336.25) (556.61, 343.26) (544.49, 343.26)      /T1_2 the	<|special_separator|>
-(306.13, 325.50) (334.65, 325.50) (334.65, 332.51) (306.13, 332.51)      /T1_2 article's	<|special_separator|>
-(336.41, 325.50) (367.45, 325.50) (367.45, 332.51) (336.41, 332.51)      /T1_2 Creative	<|special_separator|>
-(369.21, 325.50) (438.94, 325.50) (438.94, 332.51) (369.21, 332.51)      /T1_2 Commonslicence,	<|special_separator|>
-(440.70, 325.50) (464.73, 325.50) (464.73, 332.51) (440.70, 332.51)      /T1_2 unless	<|special_separator|>
-(466.49, 325.50) (501.46, 325.50) (501.46, 332.51) (466.49, 332.51)      /T1_2 indicated	<|special_separator|>
-(503.22, 325.50) (539.85, 325.50) (539.85, 332.51) (503.22, 332.51)      /T1_2 otherwise	<|special_separator|>
-(541.61, 325.50) (548.41, 325.50) (548.41, 332.51) (541.61, 332.51)      /T1_2 in	<|special_separator|>
-(550.17, 325.50) (554.49, 325.50) (554.49, 332.51) (550.17, 332.51)      /T1_2 a	<|special_separator|>
-(306.13, 314.75) (327.90, 314.75) (327.90, 321.76) (306.13, 321.76)      /T1_2 credit	<|special_separator|>
-(329.66, 314.75) (343.30, 314.75) (343.30, 321.76) (329.66, 321.76)      /T1_2 line	<|special_separator|>
-(345.06, 314.75) (352.71, 314.75) (352.71, 321.76) (345.06, 321.76)      /T1_2 to	<|special_separator|>
-(354.47, 314.75) (366.59, 314.75) (366.59, 321.76) (354.47, 321.76)      /T1_2 the	<|special_separator|>
-(368.35, 314.75) (401.11, 314.75) (401.11, 321.76) (368.35, 321.76)      /T1_2 material.	<|special_separator|>
-(402.87, 314.75) (407.83, 314.75) (407.83, 321.76) (402.87, 321.76)      /T1_2 If	<|special_separator|>
-(409.59, 314.75) (440.12, 314.75) (440.12, 321.76) (409.59, 321.76)      /T1_2 material	<|special_separator|>
-(441.88, 314.75) (447.74, 314.75) (447.74, 321.76) (441.88, 321.76)      /T1_2 is	<|special_separator|>
-(449.50, 314.75) (461.90, 314.75) (461.90, 321.76) (449.50, 321.76)      /T1_2 not	<|special_separator|>
-(463.66, 314.75) (496.55, 314.75) (496.55, 321.76) (463.66, 321.76)      /T1_2 included	<|special_separator|>
-(498.31, 314.75) (505.11, 314.75) (505.11, 321.76) (498.31, 321.76)      /T1_2 in	<|special_separator|>
-(506.87, 314.75) (518.99, 314.75) (518.99, 321.76) (506.87, 321.76)      /T1_2 the	<|special_separator|>
-(520.75, 314.75) (549.26, 314.75) (549.26, 321.76) (520.75, 321.76)      /T1_2 article's	<|special_separator|>
-(306.13, 303.99) (337.18, 303.99) (337.18, 311.01) (306.13, 311.01)      /T1_2 Creative	<|special_separator|>
-(338.94, 303.99) (406.48, 303.99) (406.48, 311.01) (338.94, 311.01)      /T1_2 Commonslicence	<|special_separator|>
-(408.24, 303.99) (422.31, 303.99) (422.31, 311.01) (408.24, 311.01)      /T1_2 and	<|special_separator|>
-(424.07, 303.99) (440.70, 303.99) (440.70, 311.01) (424.07, 311.01)      /T1_2 your	<|special_separator|>
-(442.46, 303.99) (475.86, 303.99) (475.86, 311.01) (442.46, 311.01)      /T1_2 intended	<|special_separator|>
-(477.62, 303.99) (490.71, 303.99) (490.71, 311.01) (477.62, 311.01)      /T1_2 use	<|special_separator|>
-(492.47, 303.99) (498.33, 303.99) (498.33, 311.01) (492.47, 311.01)      /T1_2 is	<|special_separator|>
-(500.09, 303.99) (512.50, 303.99) (512.50, 311.01) (500.09, 311.01)      /T1_2 not	<|special_separator|>
-(514.26, 303.99) (551.19, 303.99) (551.19, 311.01) (514.26, 311.01)      /T1_2 permitted	<|special_separator|>
-(306.13, 293.24) (315.28, 293.24) (315.28, 300.26) (306.13, 300.26)      /T1_2 by	<|special_separator|>
-(317.04, 293.24) (350.44, 293.24) (350.44, 300.26) (317.04, 300.26)      /T1_2 statutory	<|special_separator|>
-(352.20, 293.24) (390.34, 293.24) (390.34, 300.26) (352.20, 300.26)      /T1_2 regulation	<|special_separator|>
-(392.10, 293.24) (399.92, 293.24) (399.92, 300.26) (392.10, 300.26)      /T1_2 or	<|special_separator|>
-(401.68, 293.24) (432.49, 293.24) (432.49, 300.26) (401.68, 300.26)      /T1_2 exceeds	<|special_separator|>
-(434.25, 293.24) (446.37, 293.24) (446.37, 300.26) (434.25, 300.26)      /T1_2 the	<|special_separator|>
-(448.13, 293.24) (485.07, 293.24) (485.07, 300.26) (448.13, 300.26)      /T1_2 permitted	<|special_separator|>
-(486.83, 293.24) (502.11, 293.24) (502.11, 300.26) (486.83, 300.26)      /T1_2 use,	<|special_separator|>
-(503.87, 293.24) (517.52, 293.24) (517.52, 300.26) (503.87, 300.26)      /T1_2 you	<|special_separator|>
-(519.28, 293.24) (532.29, 293.24) (532.29, 300.26) (519.28, 300.26)      /T1_2 will	<|special_separator|>
-(534.05, 293.24) (552.87, 293.24) (552.87, 300.26) (534.05, 300.26)      /T1_2 need	<|special_separator|>
-(306.13, 282.49) (313.78, 282.49) (313.78, 289.50) (306.13, 289.50)      /T1_2 to	<|special_separator|>
-(315.54, 282.49) (339.30, 282.49) (339.30, 289.50) (315.54, 289.50)      /T1_2 obtain	<|special_separator|>
-(341.06, 282.49) (382.06, 282.49) (382.06, 289.50) (341.06, 289.50)      /T1_2 permission	<|special_separator|>
-(383.82, 282.49) (412.12, 282.49) (412.12, 289.50) (383.82, 289.50)      /T1_2 directly	<|special_separator|>
-(413.88, 282.49) (431.58, 282.49) (431.58, 289.50) (413.88, 289.50)      /T1_2 from	<|special_separator|>
-(433.34, 282.49) (445.46, 282.49) (445.46, 289.50) (433.34, 289.50)      /T1_2 the	<|special_separator|>
-(447.22, 282.49) (483.28, 282.49) (483.28, 289.50) (447.22, 289.50)      /T1_2 copyright	<|special_separator|>
-(485.04, 282.49) (511.20, 282.49) (511.20, 289.50) (485.04, 289.50)      /T1_2 holder.	<|special_separator|>
-(512.96, 282.49) (521.66, 282.49) (521.66, 289.50) (512.96, 289.50)      /T1_2 To	<|special_separator|>
-(523.42, 282.49) (540.52, 282.49) (540.52, 289.50) (523.42, 289.50)      /T1_2 view	<|special_separator|>
-(542.28, 282.49) (546.60, 282.49) (546.60, 289.50) (542.28, 289.50)      /T1_2 a	<|special_separator|>
-(306.13, 271.74) (324.62, 271.74) (324.62, 278.75) (306.13, 278.75)      /T1_2 copy	<|special_separator|>
-(326.38, 271.74) (333.96, 271.74) (333.96, 278.75) (326.38, 278.75)      /T1_2 of	<|special_separator|>
-(335.72, 271.74) (349.15, 271.74) (349.15, 278.75) (335.72, 278.75)      /T1_2 this	<|special_separator|>
-(350.91, 271.74) (380.34, 271.74) (380.34, 278.75) (350.91, 278.75)      /T1_2 licence,	<|special_separator|>
-(382.10, 271.74) (397.04, 271.74) (397.04, 278.75) (382.10, 278.75)      /T1_2 visit	<|special_separator|>
-(398.80, 271.74) (540.00, 271.74) (540.00, 278.75) (398.80, 278.75)      /T1_2 http://creativecommons.org/licenses/	<|special_separator|>
-(306.13, 260.99) (359.08, 260.99) (359.08, 268.00) (306.13, 268.00)      /T1_2 by-nc-nd/4.0/.	<|special_separator|>
-(306.13, 239.48) (363.27, 239.48) (363.27, 246.50) (306.13, 246.50)      /T1_2 ©TheAuthor(s)	<|special_separator|>
-(364.87, 239.48) (383.70, 239.48) (383.70, 246.50) (364.87, 246.50)      /T1_2 2026

groundtruth/parser/font_11.pdf.page_no_14.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8d661d45a7234c7c58d486bcce8f7be101c90555cfe3f12370230f2020492319
-size 1570362

groundtruth/parser/font_11.pdf.page_no_14.py.json.char.txt

@@ -1,271 +0,0 @@
-(039.69, 020.40) (045.65, 020.40) (045.65, 027.41) (039.69, 027.41)      /T1_0 N	<|special_separator|>
-(045.53, 020.40) (049.97, 020.40) (049.97, 027.41) (045.53, 027.41)      /T1_0 a	<|special_separator|>
-(049.89, 020.40) (052.84, 020.40) (052.84, 027.41) (049.89, 027.41)      /T1_0 t	<|special_separator|>
-(052.76, 020.40) (057.60, 020.40) (057.60, 027.41) (052.76, 027.41)      /T1_0 u	<|special_separator|>
-(057.52, 020.40) (060.69, 020.40) (060.69, 027.41) (057.52, 027.41)      /T1_0 r	<|special_separator|>
-(060.53, 020.40) (065.17, 020.40) (065.17, 027.41) (060.53, 027.41)      /T1_0 e	<|special_separator|>
-(065.09, 020.40) (067.00, 020.40) (067.00, 027.41) (065.09, 027.41)      /T1_0  	<|special_separator|>
-(066.63, 020.40) (073.75, 020.40) (073.75, 027.41) (066.63, 027.41)      /T1_0 M	<|special_separator|>
-(073.63, 020.40) (078.07, 020.40) (078.07, 027.41) (073.63, 027.41)      /T1_0 a	<|special_separator|>
-(077.99, 020.40) (082.59, 020.40) (082.59, 027.41) (077.99, 027.41)      /T1_0 c	<|special_separator|>
-(082.51, 020.40) (087.37, 020.40) (087.37, 027.41) (082.51, 027.41)      /T1_0 h	<|special_separator|>
-(087.29, 020.40) (089.46, 020.40) (089.46, 027.41) (087.29, 027.41)      /T1_0 i	<|special_separator|>
-(089.38, 020.40) (094.25, 020.40) (094.25, 027.41) (089.38, 027.41)      /T1_0 n	<|special_separator|>
-(094.17, 020.40) (098.81, 020.40) (098.81, 027.41) (094.17, 027.41)      /T1_0 e	<|special_separator|>
-(098.73, 020.40) (100.63, 020.40) (100.63, 027.41) (098.73, 027.41)      /T1_0  	<|special_separator|>
-(100.26, 020.40) (102.72, 020.40) (102.72, 027.41) (100.26, 027.41)      /T1_0 I	<|special_separator|>
-(102.64, 020.40) (107.50, 020.40) (107.50, 027.41) (102.64, 027.41)      /T1_0 n	<|special_separator|>
-(107.42, 020.40) (110.37, 020.40) (110.37, 027.41) (107.42, 027.41)      /T1_0 t	<|special_separator|>
-(110.29, 020.40) (114.93, 020.40) (114.93, 027.41) (110.29, 027.41)      /T1_0 e	<|special_separator|>
-(114.85, 020.40) (117.36, 020.40) (117.36, 027.41) (114.85, 027.41)      /T1_0 l	<|special_separator|>
-(117.28, 020.40) (119.78, 020.40) (119.78, 027.41) (117.28, 027.41)      /T1_0 l	<|special_separator|>
-(119.70, 020.40) (121.87, 020.40) (121.87, 027.41) (119.70, 027.41)      /T1_0 i	<|special_separator|>
-(121.79, 020.40) (126.85, 020.40) (126.85, 027.41) (121.79, 027.41)      /T1_0 g	<|special_separator|>
-(126.77, 020.40) (131.41, 020.40) (131.41, 027.41) (126.77, 027.41)      /T1_0 e	<|special_separator|>
-(131.33, 020.40) (136.20, 020.40) (136.20, 027.41) (131.33, 027.41)      /T1_0 n	<|special_separator|>
-(136.12, 020.40) (140.72, 020.40) (140.72, 027.41) (136.12, 027.41)      /T1_0 c	<|special_separator|>
-(140.64, 020.40) (145.28, 020.40) (145.28, 027.41) (140.64, 027.41)      /T1_0 e	<|special_separator|>
-(039.69, 757.67) (046.72, 757.67) (046.72, 766.42) (039.69, 766.42)      /T1_0 A	<|special_separator|>
-(046.62, 757.67) (050.59, 757.67) (050.59, 766.42) (046.62, 766.42)      /T1_0 r	<|special_separator|>
-(050.54, 757.67) (054.23, 757.67) (054.23, 766.42) (050.54, 766.42)      /T1_0 t	<|special_separator|>
-(054.13, 757.67) (056.84, 757.67) (056.84, 766.42) (054.13, 766.42)      /T1_0 i	<|special_separator|>
-(056.74, 757.67) (062.49, 757.67) (062.49, 766.42) (056.74, 766.42)      /T1_0 c	<|special_separator|>
-(062.39, 757.67) (065.52, 757.67) (065.52, 766.42) (062.39, 766.42)      /T1_0 l	<|special_separator|>
-(065.42, 757.67) (071.22, 757.67) (071.22, 766.42) (065.42, 766.42)      /T1_0 e	<|special_separator|>
-(391.20, 757.78) (396.23, 757.78) (396.23, 765.05) (391.20, 765.05)      /T1_1 h	<|special_separator|>
-(396.23, 757.78) (399.60, 757.78) (399.60, 765.05) (396.23, 765.05)      /T1_1 t	<|special_separator|>
-(399.60, 757.78) (402.97, 757.78) (402.97, 765.05) (399.60, 765.05)      /T1_1 t	<|special_separator|>
-(402.97, 757.78) (408.09, 757.78) (408.09, 765.05) (402.97, 765.05)      /T1_1 p	<|special_separator|>
-(408.09, 757.78) (411.77, 757.78) (411.77, 765.05) (408.09, 765.05)      /T1_1 s	<|special_separator|>
-(411.77, 757.78) (414.20, 757.78) (414.20, 765.05) (411.77, 765.05)      /T1_1 :	<|special_separator|>
-(414.20, 757.78) (417.33, 757.78) (417.33, 765.05) (414.20, 765.05)      /T1_1 /	<|special_separator|>
-(416.57, 757.78) (419.69, 757.78) (419.69, 765.05) (416.57, 765.05)      /T1_1 /	<|special_separator|>
-(419.21, 757.78) (424.29, 757.78) (424.29, 765.05) (419.21, 765.05)      /T1_1 d	<|special_separator|>
-(424.29, 757.78) (429.25, 757.78) (429.25, 765.05) (424.29, 765.05)      /T1_1 o	<|special_separator|>
-(429.25, 757.78) (431.73, 757.78) (431.73, 765.05) (429.25, 765.05)      /T1_1 i	<|special_separator|>
-(431.73, 757.78) (434.15, 757.78) (434.15, 765.05) (431.73, 765.05)      /T1_1 .	<|special_separator|>
-(433.95, 757.78) (438.90, 757.78) (438.90, 765.05) (433.95, 765.05)      /T1_1 o	<|special_separator|>
-(438.90, 757.78) (442.49, 757.78) (442.49, 765.05) (438.90, 765.05)      /T1_1 r	<|special_separator|>
-(442.41, 757.78) (447.19, 757.78) (447.19, 765.05) (442.41, 765.05)      /T1_1 g	<|special_separator|>
-(447.23, 757.78) (450.36, 757.78) (450.36, 765.05) (447.23, 765.05)      /T1_1 /	<|special_separator|>
-(450.32, 757.78) (453.95, 757.78) (453.95, 765.05) (450.32, 765.05)      /T1_1 1	<|special_separator|>
-(453.95, 757.78) (459.37, 757.78) (459.37, 765.05) (453.95, 765.05)      /T1_1 0	<|special_separator|>
-(458.97, 757.78) (461.40, 757.78) (461.40, 765.05) (458.97, 765.05)      /T1_1 .	<|special_separator|>
-(460.80, 757.78) (464.43, 757.78) (464.43, 765.05) (460.80, 765.05)      /T1_1 1	<|special_separator|>
-(464.43, 757.78) (469.85, 757.78) (469.85, 765.05) (464.43, 765.05)      /T1_1 0	<|special_separator|>
-(469.81, 757.78) (474.29, 757.78) (474.29, 765.05) (469.81, 765.05)      /T1_1 3	<|special_separator|>
-(474.29, 757.78) (479.13, 757.78) (479.13, 765.05) (474.29, 765.05)      /T1_1 8	<|special_separator|>
-(478.93, 757.78) (482.06, 757.78) (482.06, 765.05) (478.93, 765.05)      /T1_1 /	<|special_separator|>
-(481.62, 757.78) (485.31, 757.78) (485.31, 765.05) (481.62, 765.05)      /T1_1 s	<|special_separator|>
-(485.31, 757.78) (490.21, 757.78) (490.21, 765.05) (485.31, 765.05)      /T1_1 4	<|special_separator|>
-(490.21, 757.78) (494.73, 757.78) (494.73, 765.05) (490.21, 765.05)      /T1_1 2	<|special_separator|>
-(494.76, 757.78) (499.28, 757.78) (499.28, 765.05) (494.76, 765.05)      /T1_1 2	<|special_separator|>
-(499.28, 757.78) (503.60, 757.78) (503.60, 765.05) (499.28, 765.05)      /T1_1 5	<|special_separator|>
-(503.52, 757.78) (508.44, 757.78) (508.44, 765.05) (503.52, 765.05)      /T1_1 6	<|special_separator|>
-(508.40, 757.78) (510.92, 757.78) (510.92, 765.05) (508.40, 765.05)      /T1_1 -	<|special_separator|>
-(510.92, 757.78) (516.34, 757.78) (516.34, 765.05) (510.92, 765.05)      /T1_1 0	<|special_separator|>
-(516.34, 757.78) (520.85, 757.78) (520.85, 765.05) (516.34, 765.05)      /T1_1 2	<|special_separator|>
-(520.81, 757.78) (525.73, 757.78) (525.73, 765.05) (520.81, 765.05)      /T1_1 6	<|special_separator|>
-(525.69, 757.78) (528.20, 757.78) (528.20, 765.05) (525.69, 765.05)      /T1_1 -	<|special_separator|>
-(528.20, 757.78) (533.63, 757.78) (533.63, 765.05) (528.20, 765.05)      /T1_1 0	<|special_separator|>
-(533.63, 757.78) (537.26, 757.78) (537.26, 765.05) (533.63, 765.05)      /T1_1 1	<|special_separator|>
-(537.26, 757.78) (541.77, 757.78) (541.77, 765.05) (537.26, 765.05)      /T1_1 2	<|special_separator|>
-(541.57, 757.78) (546.48, 757.78) (546.48, 765.05) (541.57, 765.05)      /T1_1 4	<|special_separator|>
-(546.48, 757.78) (550.99, 757.78) (550.99, 765.05) (546.48, 765.05)      /T1_1 2	<|special_separator|>
-(550.91, 757.78) (553.42, 757.78) (553.42, 765.05) (550.91, 765.05)      /T1_1 -	<|special_separator|>
-(553.42, 757.78) (558.27, 757.78) (558.27, 765.05) (553.42, 765.05)      /T1_1 8	<|special_separator|>
-(039.69, 343.17) (043.70, 343.17) (043.70, 349.53) (039.69, 349.53)      /T1_1 E	<|special_separator|>
-(043.68, 343.17) (047.81, 343.17) (047.81, 349.53) (043.68, 349.53)      /T1_1 x	<|special_separator|>
-(047.65, 343.17) (050.60, 343.17) (050.60, 349.53) (047.65, 349.53)      /T1_1 t	<|special_separator|>
-(050.55, 343.17) (054.36, 343.17) (054.36, 349.53) (050.55, 349.53)      /T1_1 e	<|special_separator|>
-(054.34, 343.17) (058.74, 343.17) (058.74, 349.53) (054.34, 349.53)      /T1_1 n	<|special_separator|>
-(058.72, 343.17) (063.17, 343.17) (063.17, 349.53) (058.72, 349.53)      /T1_1 d	<|special_separator|>
-(063.15, 343.17) (066.96, 343.17) (066.96, 349.53) (063.15, 349.53)      /T1_1 e	<|special_separator|>
-(066.94, 343.17) (071.39, 343.17) (071.39, 349.53) (066.94, 349.53)      /T1_1 d	<|special_separator|>
-(071.37, 343.17) (072.54, 343.17) (072.54, 349.53) (071.37, 349.53)      /T1_1  	<|special_separator|>
-(072.35, 343.17) (077.63, 343.17) (077.63, 349.53) (072.35, 349.53)      /T1_1 D	<|special_separator|>
-(077.62, 343.17) (081.49, 343.17) (081.49, 349.53) (077.62, 349.53)      /T1_1 a	<|special_separator|>
-(081.47, 343.17) (084.42, 343.17) (084.42, 349.53) (081.47, 349.53)      /T1_1 t	<|special_separator|>
-(084.44, 343.17) (088.31, 343.17) (088.31, 349.53) (084.44, 349.53)      /T1_1 a	<|special_separator|>
-(088.30, 343.17) (089.47, 343.17) (089.47, 349.53) (088.30, 349.53)      /T1_1  	<|special_separator|>
-(089.28, 343.17) (093.33, 343.17) (093.33, 349.53) (089.28, 349.53)      /T1_1 F	<|special_separator|>
-(093.24, 343.17) (095.41, 343.17) (095.41, 349.53) (093.24, 349.53)      /T1_1 i	<|special_separator|>
-(095.40, 343.17) (099.59, 343.17) (099.59, 349.53) (095.40, 349.53)      /T1_1 g	<|special_separator|>
-(099.57, 343.17) (101.69, 343.17) (101.69, 349.53) (099.57, 349.53)      /T1_1 .	<|special_separator|>
-(101.68, 343.17) (102.85, 343.17) (102.85, 349.53) (101.68, 349.53)      /T1_1  	<|special_separator|>
-(102.66, 343.17) (105.84, 343.17) (105.84, 349.53) (102.66, 349.53)      /T1_1 1	<|special_separator|>
-(105.82, 343.17) (106.99, 343.17) (106.99, 349.53) (105.82, 349.53)      /T1_1  	<|special_separator|>
-(106.80, 343.17) (108.48, 343.17) (108.48, 349.53) (106.80, 349.53)      /T1_1 |	<|special_separator|>
-(108.46, 343.17) (109.63, 343.17) (109.63, 349.53) (108.46, 349.53)      /T1_1  	<|special_separator|>
-(109.44, 343.17) (113.45, 343.17) (113.45, 349.53) (109.44, 349.53)      /T1_1 E	<|special_separator|>
-(113.41, 343.17) (117.29, 343.17) (117.29, 349.53) (113.41, 349.53)      /T1_1 v	<|special_separator|>
-(117.17, 343.17) (121.05, 343.17) (121.05, 349.53) (117.17, 349.53)      /T1_1 a	<|special_separator|>
-(121.03, 343.17) (123.24, 343.17) (123.24, 349.53) (121.03, 349.53)      /T1_1 l	<|special_separator|>
-(123.22, 343.17) (127.63, 343.17) (127.63, 349.53) (123.22, 349.53)      /T1_1 u	<|special_separator|>
-(127.61, 343.17) (131.48, 343.17) (131.48, 349.53) (127.61, 349.53)      /T1_1 a	<|special_separator|>
-(131.47, 343.17) (134.42, 343.17) (134.42, 349.53) (131.47, 349.53)      /T1_1 t	<|special_separator|>
-(134.40, 343.17) (136.57, 343.17) (136.57, 349.53) (134.40, 349.53)      /T1_1 i	<|special_separator|>
-(136.56, 343.17) (140.89, 343.17) (140.89, 349.53) (136.56, 349.53)      /T1_1 o	<|special_separator|>
-(140.88, 343.17) (145.27, 343.17) (145.27, 349.53) (140.88, 349.53)      /T1_1 n	<|special_separator|>
-(145.26, 343.17) (146.43, 343.17) (146.43, 349.53) (145.26, 349.53)      /T1_1  	<|special_separator|>
-(146.24, 343.17) (150.57, 343.17) (150.57, 349.53) (146.24, 349.53)      /T1_1 o	<|special_separator|>
-(150.56, 343.17) (153.12, 343.17) (153.12, 349.53) (150.56, 349.53)      /T1_1 f	<|special_separator|>
-(153.11, 343.17) (154.27, 343.17) (154.27, 349.53) (153.11, 349.53)      /T1_1  	<|special_separator|>
-(154.09, 343.17) (158.57, 343.17) (158.57, 349.53) (154.09, 349.53)      /T1_1 p	<|special_separator|>
-(158.55, 343.17) (162.35, 343.17) (162.35, 349.53) (158.55, 349.53)      /T1_1 e	<|special_separator|>
-(162.34, 343.17) (165.47, 343.17) (165.47, 349.53) (162.34, 349.53)      /T1_1 r	<|special_separator|>
-(165.53, 343.17) (168.48, 343.17) (168.48, 349.53) (165.53, 349.53)      /T1_1 t	<|special_separator|>
-(168.47, 343.17) (172.87, 343.17) (172.87, 349.53) (168.47, 349.53)      /T1_1 u	<|special_separator|>
-(172.85, 343.17) (175.99, 343.17) (175.99, 349.53) (172.85, 349.53)      /T1_1 r	<|special_separator|>
-(175.98, 343.17) (180.40, 343.17) (180.40, 349.53) (175.98, 349.53)      /T1_1 b	<|special_separator|>
-(180.35, 343.17) (184.23, 343.17) (184.23, 349.53) (180.35, 349.53)      /T1_1 a	<|special_separator|>
-(184.21, 343.17) (187.16, 343.17) (187.16, 349.53) (184.21, 349.53)      /T1_1 t	<|special_separator|>
-(187.14, 343.17) (189.31, 343.17) (189.31, 349.53) (187.14, 349.53)      /T1_1 i	<|special_separator|>
-(189.30, 343.17) (193.63, 343.17) (193.63, 349.53) (189.30, 349.53)      /T1_1 o	<|special_separator|>
-(193.62, 343.17) (198.02, 343.17) (198.02, 349.53) (193.62, 349.53)      /T1_1 n	<|special_separator|>
-(198.00, 343.17) (199.17, 343.17) (199.17, 349.53) (198.00, 349.53)      /T1_1  	<|special_separator|>
-(198.98, 343.17) (203.46, 343.17) (203.46, 349.53) (198.98, 349.53)      /T1_1 p	<|special_separator|>
-(203.45, 343.17) (206.58, 343.17) (206.58, 349.53) (203.45, 349.53)      /T1_1 r	<|special_separator|>
-(206.54, 343.17) (210.34, 343.17) (210.34, 349.53) (206.54, 349.53)      /T1_1 e	<|special_separator|>
-(210.32, 343.17) (214.77, 343.17) (214.77, 349.53) (210.32, 349.53)      /T1_1 d	<|special_separator|>
-(214.76, 343.17) (216.93, 343.17) (216.93, 349.53) (214.76, 349.53)      /T1_1 i	<|special_separator|>
-(216.91, 343.17) (220.54, 343.17) (220.54, 349.53) (216.91, 349.53)      /T1_1 c	<|special_separator|>
-(220.60, 343.17) (223.54, 343.17) (223.54, 349.53) (220.60, 349.53)      /T1_1 t	<|special_separator|>
-(223.53, 343.17) (225.70, 343.17) (225.70, 349.53) (223.53, 349.53)      /T1_1 i	<|special_separator|>
-(225.68, 343.17) (230.02, 343.17) (230.02, 349.53) (225.68, 349.53)      /T1_1 o	<|special_separator|>
-(230.00, 343.17) (234.40, 343.17) (234.40, 349.53) (230.00, 349.53)      /T1_1 n	<|special_separator|>
-(234.39, 343.17) (235.55, 343.17) (235.55, 349.53) (234.39, 349.53)      /T1_1  	<|special_separator|>
-(235.37, 343.17) (239.70, 343.17) (239.70, 349.53) (235.37, 349.53)      /T1_1 o	<|special_separator|>
-(239.68, 343.17) (244.08, 343.17) (244.08, 349.53) (239.68, 349.53)      /T1_1 n	<|special_separator|>
-(244.07, 343.17) (245.24, 343.17) (245.24, 349.53) (244.07, 349.53)      /T1_1  	<|special_separator|>
-(245.05, 343.17) (247.99, 343.17) (247.99, 349.53) (245.05, 349.53)      /T1_1 t	<|special_separator|>
-(247.98, 343.17) (252.38, 343.17) (252.38, 349.53) (247.98, 349.53)      /T1_1 h	<|special_separator|>
-(252.37, 343.17) (256.17, 343.17) (256.17, 349.53) (252.37, 349.53)      /T1_1 e	<|special_separator|>
-(256.16, 343.17) (257.32, 343.17) (257.32, 349.53) (256.16, 349.53)      /T1_1  	<|special_separator|>
-(257.14, 343.17) (261.43, 343.17) (261.43, 349.53) (257.14, 349.53)      /T1_1 4	<|special_separator|>
-(261.41, 343.17) (263.58, 343.17) (263.58, 349.53) (261.41, 349.53)      /T1_1 i	<|special_separator|>
-(263.57, 343.17) (264.74, 343.17) (264.74, 349.53) (263.57, 349.53)      /T1_1  	<|special_separator|>
-(264.55, 343.17) (268.99, 343.17) (268.99, 349.53) (264.55, 349.53)      /T1_1 d	<|special_separator|>
-(268.98, 343.17) (272.85, 343.17) (272.85, 349.53) (268.98, 349.53)      /T1_1 a	<|special_separator|>
-(272.84, 343.17) (275.78, 343.17) (275.78, 349.53) (272.84, 349.53)      /T1_1 t	<|special_separator|>
-(275.81, 343.17) (279.68, 343.17) (279.68, 349.53) (275.81, 349.53)      /T1_1 a	<|special_separator|>
-(279.66, 343.17) (282.89, 343.17) (282.89, 349.53) (279.66, 349.53)      /T1_1 s	<|special_separator|>
-(282.88, 343.17) (286.68, 343.17) (286.68, 349.53) (282.88, 349.53)      /T1_1 e	<|special_separator|>
-(286.66, 343.17) (289.61, 343.17) (289.61, 349.53) (286.66, 349.53)      /T1_1 t	<|special_separator|>
-(289.60, 343.17) (291.72, 343.17) (291.72, 349.53) (289.60, 349.53)      /T1_1 .	<|special_separator|>
-(291.70, 343.04) (292.94, 343.04) (292.94, 349.53) (291.70, 349.53)      /T1_2  	<|special_separator|>
-(292.74, 343.04) (296.60, 343.04) (296.60, 349.53) (292.74, 349.53)      /T1_2 E	<|special_separator|>
-(296.59, 343.04) (300.22, 343.04) (300.22, 349.53) (296.59, 349.53)      /T1_2 a	<|special_separator|>
-(300.20, 343.04) (303.79, 343.04) (303.79, 349.53) (300.20, 349.53)      /T1_2 c	<|special_separator|>
-(303.77, 343.04) (307.96, 343.04) (307.96, 349.53) (303.77, 349.53)      /T1_2 h	<|special_separator|>
-(307.94, 343.04) (309.18, 343.04) (309.18, 349.53) (307.94, 349.53)      /T1_2  	<|special_separator|>
-(308.98, 343.04) (313.31, 343.04) (313.31, 349.53) (308.98, 349.53)      /T1_2 p	<|special_separator|>
-(313.29, 343.04) (317.53, 343.04) (317.53, 349.53) (313.29, 349.53)      /T1_2 o	<|special_separator|>
-(317.51, 343.04) (319.51, 343.04) (319.51, 349.53) (317.51, 349.53)      /T1_2 i	<|special_separator|>
-(319.50, 343.04) (323.68, 343.04) (323.68, 349.53) (319.50, 349.53)      /T1_2 n	<|special_separator|>
-(323.67, 343.04) (326.35, 343.04) (326.35, 349.53) (323.67, 349.53)      /T1_2 t	<|special_separator|>
-(326.33, 343.04) (327.56, 343.04) (327.56, 349.53) (326.33, 349.53)      /T1_2  	<|special_separator|>
-(327.37, 343.04) (330.95, 343.04) (330.95, 349.53) (327.37, 349.53)      /T1_2 c	<|special_separator|>
-(330.90, 343.04) (335.14, 343.04) (335.14, 349.53) (330.90, 349.53)      /T1_2 o	<|special_separator|>
-(335.13, 343.04) (338.03, 343.04) (338.03, 349.53) (335.13, 349.53)      /T1_2 r	<|special_separator|>
-(338.02, 343.04) (340.92, 343.04) (340.92, 349.53) (338.02, 349.53)      /T1_2 r	<|special_separator|>
-(340.80, 343.04) (344.53, 343.04) (344.53, 349.53) (340.80, 349.53)      /T1_2 e	<|special_separator|>
-(344.51, 343.04) (347.69, 343.04) (347.69, 349.53) (344.51, 349.53)      /T1_2 s	<|special_separator|>
-(347.67, 343.04) (352.00, 343.04) (352.00, 349.53) (347.67, 349.53)      /T1_2 p	<|special_separator|>
-(351.98, 343.04) (356.22, 343.04) (356.22, 349.53) (351.98, 349.53)      /T1_2 o	<|special_separator|>
-(356.20, 343.04) (360.38, 343.04) (360.38, 349.53) (356.20, 349.53)      /T1_2 n	<|special_separator|>
-(360.37, 343.04) (364.67, 343.04) (364.67, 349.53) (360.37, 349.53)      /T1_2 d	<|special_separator|>
-(364.66, 343.04) (367.83, 343.04) (367.83, 349.53) (364.66, 349.53)      /T1_2 s	<|special_separator|>
-(367.82, 343.04) (369.05, 343.04) (369.05, 349.53) (367.82, 349.53)      /T1_2  	<|special_separator|>
-(368.85, 343.04) (371.53, 343.04) (371.53, 349.53) (368.85, 349.53)      /T1_2 t	<|special_separator|>
-(371.49, 343.04) (375.72, 343.04) (375.72, 349.53) (371.49, 349.53)      /T1_2 o	<|special_separator|>
-(375.71, 343.04) (376.94, 343.04) (376.94, 349.53) (375.71, 349.53)      /T1_2  	<|special_separator|>
-(376.74, 343.04) (380.37, 343.04) (380.37, 349.53) (376.74, 349.53)      /T1_2 a	<|special_separator|>
-(380.36, 343.04) (381.59, 343.04) (381.59, 349.53) (380.36, 349.53)      /T1_2  	<|special_separator|>
-(381.39, 343.04) (387.73, 343.04) (387.73, 349.53) (381.39, 349.53)      /T1_2 m	<|special_separator|>
-(387.71, 343.04) (391.95, 343.04) (391.95, 349.53) (387.71, 349.53)      /T1_2 o	<|special_separator|>
-(391.93, 343.04) (396.24, 343.04) (396.24, 349.53) (391.93, 349.53)      /T1_2 d	<|special_separator|>
-(396.23, 343.04) (399.95, 343.04) (399.95, 349.53) (396.23, 349.53)      /T1_2 e	<|special_separator|>
-(399.94, 343.04) (401.93, 343.04) (401.93, 349.53) (399.94, 349.53)      /T1_2 l	<|special_separator|>
-(401.92, 343.04) (403.15, 343.04) (403.15, 349.53) (401.92, 349.53)      /T1_2  	<|special_separator|>
-(402.95, 343.04) (405.63, 343.04) (405.63, 349.53) (402.95, 349.53)      /T1_2 t	<|special_separator|>
-(405.62, 343.04) (408.52, 343.04) (408.52, 349.53) (405.62, 349.53)      /T1_2 r	<|special_separator|>
-(408.44, 343.04) (412.07, 343.04) (412.07, 349.53) (408.44, 349.53)      /T1_2 a	<|special_separator|>
-(412.05, 343.04) (414.06, 343.04) (414.06, 349.53) (412.05, 349.53)      /T1_2 i	<|special_separator|>
-(414.04, 343.04) (418.22, 343.04) (418.22, 349.53) (414.04, 349.53)      /T1_2 n	<|special_separator|>
-(418.21, 343.04) (421.93, 343.04) (421.93, 349.53) (418.21, 349.53)      /T1_2 e	<|special_separator|>
-(421.92, 343.04) (426.22, 343.04) (426.22, 349.53) (421.92, 349.53)      /T1_2 d	<|special_separator|>
-(426.21, 343.04) (427.44, 343.04) (427.44, 349.53) (426.21, 349.53)      /T1_2  	<|special_separator|>
-(427.24, 343.04) (431.48, 343.04) (431.48, 349.53) (427.24, 349.53)      /T1_2 o	<|special_separator|>
-(431.46, 343.04) (435.64, 343.04) (435.64, 349.53) (431.46, 349.53)      /T1_2 n	<|special_separator|>
-(435.63, 343.04) (436.86, 343.04) (436.86, 349.53) (435.63, 349.53)      /T1_2  	<|special_separator|>
-(436.67, 343.04) (440.90, 343.04) (440.90, 349.53) (436.67, 349.53)      /T1_2 o	<|special_separator|>
-(440.89, 343.04) (445.07, 343.04) (445.07, 349.53) (440.89, 349.53)      /T1_2 n	<|special_separator|>
-(445.05, 343.04) (448.78, 343.04) (448.78, 349.53) (445.05, 349.53)      /T1_2 e	<|special_separator|>
-(448.76, 343.04) (449.99, 343.04) (449.99, 349.53) (448.76, 349.53)      /T1_2  	<|special_separator|>
-(449.80, 343.04) (454.03, 343.04) (454.03, 349.53) (449.80, 349.53)      /T1_2 o	<|special_separator|>
-(453.99, 343.04) (456.51, 343.04) (456.51, 349.53) (453.99, 349.53)      /T1_2 f	<|special_separator|>
-(456.50, 343.04) (457.73, 343.04) (457.73, 349.53) (456.50, 349.53)      /T1_2  	<|special_separator|>
-(457.53, 343.04) (461.36, 343.04) (461.36, 349.53) (457.53, 349.53)      /T1_2 3	<|special_separator|>
-(461.28, 343.04) (464.87, 343.04) (464.87, 349.53) (461.28, 349.53)      /T1_2 5	<|special_separator|>
-(464.85, 343.04) (466.08, 343.04) (466.08, 349.53) (464.85, 349.53)      /T1_2  	<|special_separator|>
-(465.89, 343.04) (468.57, 343.04) (468.57, 349.53) (465.89, 349.53)      /T1_2 t	<|special_separator|>
-(468.55, 343.04) (471.46, 343.04) (471.46, 349.53) (468.55, 349.53)      /T1_2 r	<|special_separator|>
-(471.34, 343.04) (475.07, 343.04) (475.07, 349.53) (471.34, 349.53)      /T1_2 e	<|special_separator|>
-(475.05, 343.04) (478.68, 343.04) (478.68, 349.53) (475.05, 349.53)      /T1_2 a	<|special_separator|>
-(478.66, 343.04) (481.34, 343.04) (481.34, 349.53) (478.66, 349.53)      /T1_2 t	<|special_separator|>
-(481.33, 343.04) (487.67, 343.04) (487.67, 349.53) (481.33, 349.53)      /T1_2 m	<|special_separator|>
-(487.65, 343.04) (491.38, 343.04) (491.38, 349.53) (487.65, 349.53)      /T1_2 e	<|special_separator|>
-(491.36, 343.04) (495.54, 343.04) (495.54, 349.53) (491.36, 349.53)      /T1_2 n	<|special_separator|>
-(495.53, 343.04) (498.21, 343.04) (498.21, 349.53) (495.53, 349.53)      /T1_2 t	<|special_separator|>
-(498.19, 343.04) (501.36, 343.04) (501.36, 349.53) (498.19, 349.53)      /T1_2 s	<|special_separator|>
-(501.35, 343.04) (503.24, 343.04) (503.24, 349.53) (501.35, 349.53)      /T1_2 .	<|special_separator|>
-(503.23, 343.04) (504.46, 343.04) (504.46, 349.53) (503.23, 349.53)      /T1_2  	<|special_separator|>
-(504.26, 343.04) (508.13, 343.04) (508.13, 349.53) (504.26, 349.53)      /T1_2 E	<|special_separator|>
-(508.11, 343.04) (511.02, 343.04) (511.02, 349.53) (508.11, 349.53)      /T1_2 r	<|special_separator|>
-(511.00, 343.04) (513.91, 343.04) (513.91, 349.53) (511.00, 349.53)      /T1_2 r	<|special_separator|>
-(513.79, 343.04) (518.03, 343.04) (518.03, 349.53) (513.79, 349.53)      /T1_2 o	<|special_separator|>
-(518.01, 343.04) (520.92, 343.04) (520.92, 349.53) (518.01, 349.53)      /T1_2 r	<|special_separator|>
-(520.90, 343.04) (522.13, 343.04) (522.13, 349.53) (520.90, 349.53)      /T1_2  	<|special_separator|>
-(521.94, 343.04) (526.19, 343.04) (526.19, 349.53) (521.94, 349.53)      /T1_2 b	<|special_separator|>
-(526.11, 343.04) (529.73, 343.04) (529.73, 349.53) (526.11, 349.53)      /T1_2 a	<|special_separator|>
-(529.72, 343.04) (532.62, 343.04) (532.62, 349.53) (529.72, 349.53)      /T1_2 r	<|special_separator|>
-(532.61, 343.04) (535.78, 343.04) (535.78, 349.53) (532.61, 349.53)      /T1_2 s	<|special_separator|>
-(535.78, 343.04) (537.01, 343.04) (537.01, 349.53) (535.78, 349.53)      /T1_2  	<|special_separator|>
-(039.69, 333.03) (041.69, 333.03) (041.69, 339.53) (039.69, 339.53)      /T1_2 i	<|special_separator|>
-(041.67, 333.03) (045.85, 333.03) (045.85, 339.53) (041.67, 339.53)      /T1_2 n	<|special_separator|>
-(045.84, 333.03) (050.14, 333.03) (050.14, 339.53) (045.84, 339.53)      /T1_2 d	<|special_separator|>
-(050.13, 333.03) (052.13, 333.03) (052.13, 339.53) (050.13, 339.53)      /T1_2 i	<|special_separator|>
-(052.12, 333.03) (055.70, 333.03) (055.70, 339.53) (052.12, 339.53)      /T1_2 c	<|special_separator|>
-(055.69, 333.03) (059.31, 333.03) (059.31, 339.53) (055.69, 339.53)      /T1_2 a	<|special_separator|>
-(059.30, 333.03) (061.98, 333.03) (061.98, 339.53) (059.30, 339.53)      /T1_2 t	<|special_separator|>
-(061.90, 333.03) (065.62, 333.03) (065.62, 339.53) (061.90, 339.53)      /T1_2 e	<|special_separator|>
-(065.61, 333.03) (066.84, 333.03) (066.84, 339.53) (065.61, 339.53)      /T1_2  	<|special_separator|>
-(066.64, 333.03) (069.32, 333.03) (069.32, 339.53) (066.64, 339.53)      /T1_2 t	<|special_separator|>
-(069.31, 333.03) (073.50, 333.03) (073.50, 339.53) (069.31, 339.53)      /T1_2 h	<|special_separator|>
-(073.48, 333.03) (077.21, 333.03) (077.21, 339.53) (073.48, 339.53)      /T1_2 e	<|special_separator|>
-(077.19, 333.03) (078.42, 333.03) (078.42, 339.53) (077.19, 339.53)      /T1_2  	<|special_separator|>
-(078.23, 333.03) (082.41, 333.03) (082.41, 339.53) (078.23, 339.53)      /T1_2 9	<|special_separator|>
-(082.15, 333.03) (085.73, 333.03) (085.73, 339.53) (082.15, 339.53)      /T1_2 5	<|special_separator|>
-(085.72, 333.03) (090.58, 333.03) (090.58, 339.53) (085.72, 339.53)      /T1_2 %	<|special_separator|>
-(090.56, 333.03) (091.80, 333.03) (091.80, 339.53) (090.56, 339.53)      /T1_2  	<|special_separator|>
-(091.60, 333.03) (095.18, 333.03) (095.18, 339.53) (091.60, 339.53)      /T1_2 c	<|special_separator|>
-(095.14, 333.03) (099.37, 333.03) (099.37, 339.53) (095.14, 339.53)      /T1_2 o	<|special_separator|>
-(099.36, 333.03) (103.54, 333.03) (103.54, 339.53) (099.36, 339.53)      /T1_2 n	<|special_separator|>
-(103.52, 333.03) (106.05, 333.03) (106.05, 339.53) (103.52, 339.53)      /T1_2 f	<|special_separator|>
-(106.11, 333.03) (108.11, 333.03) (108.11, 339.53) (106.11, 339.53)      /T1_2 i	<|special_separator|>
-(108.10, 333.03) (112.40, 333.03) (112.40, 339.53) (108.10, 339.53)      /T1_2 d	<|special_separator|>
-(112.39, 333.03) (116.11, 333.03) (116.11, 339.53) (112.39, 339.53)      /T1_2 e	<|special_separator|>
-(116.10, 333.03) (120.28, 333.03) (120.28, 339.53) (116.10, 339.53)      /T1_2 n	<|special_separator|>
-(120.26, 333.03) (123.85, 333.03) (123.85, 339.53) (120.26, 339.53)      /T1_2 c	<|special_separator|>
-(123.80, 333.03) (127.52, 333.03) (127.52, 339.53) (123.80, 339.53)      /T1_2 e	<|special_separator|>
-(127.51, 333.03) (128.74, 333.03) (128.74, 339.53) (127.51, 339.53)      /T1_2  	<|special_separator|>
-(128.54, 333.03) (130.55, 333.03) (130.55, 339.53) (128.54, 339.53)      /T1_2 i	<|special_separator|>
-(130.53, 333.03) (134.71, 333.03) (134.71, 339.53) (130.53, 339.53)      /T1_2 n	<|special_separator|>
-(134.70, 333.03) (137.38, 333.03) (137.38, 339.53) (134.70, 339.53)      /T1_2 t	<|special_separator|>
-(137.30, 333.03) (141.02, 333.03) (141.02, 339.53) (137.30, 339.53)      /T1_2 e	<|special_separator|>
-(141.01, 333.03) (143.91, 333.03) (143.91, 339.53) (141.01, 339.53)      /T1_2 r	<|special_separator|>
-(144.04, 333.03) (147.76, 333.03) (147.76, 339.53) (144.04, 339.53)      /T1_2 v	<|special_separator|>
-(147.67, 333.03) (151.30, 333.03) (151.30, 339.53) (147.67, 339.53)      /T1_2 a	<|special_separator|>
-(151.28, 333.03) (153.28, 333.03) (153.28, 339.53) (151.28, 339.53)      /T1_2 l	<|special_separator|>
-(153.27, 333.03) (155.16, 333.03) (155.16, 339.53) (153.27, 339.53)      /T1_2 .

groundtruth/parser/font_11.pdf.page_no_14.py.json.line.txt

@@ -1,6 +0,0 @@
-(039.69, 020.40) (145.28, 020.40) (145.28, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 343.17) (291.72, 343.17) (291.72, 349.53) (039.69, 349.53)      /T1_1 Extended Data Fig. 1 | Evaluation of perturbation prediction on the 4i dataset.	<|special_separator|>
-(291.70, 343.04) (537.01, 343.04) (537.01, 349.53) (291.70, 349.53)      /T1_2  Each point corresponds to a model trained on one of 35 treatments. Error bars 	<|special_separator|>
-(039.69, 333.03) (155.16, 333.03) (155.16, 339.53) (039.69, 339.53)      /T1_2 indicate the 95% confidence interval.

groundtruth/parser/font_11.pdf.page_no_14.py.json.word.txt

@@ -1,7 +0,0 @@
-(039.69, 020.40) (065.17, 020.40) (065.17, 027.41) (039.69, 027.41)      /T1_0 Nature	<|special_separator|>
-(066.63, 020.40) (145.28, 020.40) (145.28, 027.41) (066.63, 027.41)      /T1_0 MachineIntelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 343.17) (291.72, 343.17) (291.72, 349.53) (039.69, 349.53)      /T1_1 ExtendedDataFig.1|Evaluationofperturbationpredictiononthe4idataset.	<|special_separator|>
-(292.74, 343.04) (535.78, 343.04) (535.78, 349.53) (292.74, 349.53)      /T1_2 Eachpointcorrespondstoamodeltrainedononeof35treatments.Errorbars	<|special_separator|>
-(039.69, 333.03) (155.16, 333.03) (155.16, 339.53) (039.69, 339.53)      /T1_2 indicatethe95%confidenceinterval.

groundtruth/parser/font_11.pdf.page_no_15.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:36c7a7712cb4c0a9e7476219154b7092d5cc316ed77255bc8ad97da795e91e6d
-size 3316610

groundtruth/parser/font_11.pdf.page_no_15.py.json.char.txt

@@ -1,384 +0,0 @@
-(039.69, 020.40) (045.65, 020.40) (045.65, 027.41) (039.69, 027.41)      /T1_0 N	<|special_separator|>
-(045.53, 020.40) (049.97, 020.40) (049.97, 027.41) (045.53, 027.41)      /T1_0 a	<|special_separator|>
-(049.89, 020.40) (052.84, 020.40) (052.84, 027.41) (049.89, 027.41)      /T1_0 t	<|special_separator|>
-(052.76, 020.40) (057.60, 020.40) (057.60, 027.41) (052.76, 027.41)      /T1_0 u	<|special_separator|>
-(057.52, 020.40) (060.69, 020.40) (060.69, 027.41) (057.52, 027.41)      /T1_0 r	<|special_separator|>
-(060.53, 020.40) (065.17, 020.40) (065.17, 027.41) (060.53, 027.41)      /T1_0 e	<|special_separator|>
-(065.09, 020.40) (067.00, 020.40) (067.00, 027.41) (065.09, 027.41)      /T1_0  	<|special_separator|>
-(066.63, 020.40) (073.75, 020.40) (073.75, 027.41) (066.63, 027.41)      /T1_0 M	<|special_separator|>
-(073.63, 020.40) (078.07, 020.40) (078.07, 027.41) (073.63, 027.41)      /T1_0 a	<|special_separator|>
-(077.99, 020.40) (082.59, 020.40) (082.59, 027.41) (077.99, 027.41)      /T1_0 c	<|special_separator|>
-(082.51, 020.40) (087.37, 020.40) (087.37, 027.41) (082.51, 027.41)      /T1_0 h	<|special_separator|>
-(087.29, 020.40) (089.46, 020.40) (089.46, 027.41) (087.29, 027.41)      /T1_0 i	<|special_separator|>
-(089.38, 020.40) (094.25, 020.40) (094.25, 027.41) (089.38, 027.41)      /T1_0 n	<|special_separator|>
-(094.17, 020.40) (098.81, 020.40) (098.81, 027.41) (094.17, 027.41)      /T1_0 e	<|special_separator|>
-(098.73, 020.40) (100.63, 020.40) (100.63, 027.41) (098.73, 027.41)      /T1_0  	<|special_separator|>
-(100.26, 020.40) (102.72, 020.40) (102.72, 027.41) (100.26, 027.41)      /T1_0 I	<|special_separator|>
-(102.64, 020.40) (107.50, 020.40) (107.50, 027.41) (102.64, 027.41)      /T1_0 n	<|special_separator|>
-(107.42, 020.40) (110.37, 020.40) (110.37, 027.41) (107.42, 027.41)      /T1_0 t	<|special_separator|>
-(110.29, 020.40) (114.93, 020.40) (114.93, 027.41) (110.29, 027.41)      /T1_0 e	<|special_separator|>
-(114.85, 020.40) (117.36, 020.40) (117.36, 027.41) (114.85, 027.41)      /T1_0 l	<|special_separator|>
-(117.28, 020.40) (119.78, 020.40) (119.78, 027.41) (117.28, 027.41)      /T1_0 l	<|special_separator|>
-(119.70, 020.40) (121.87, 020.40) (121.87, 027.41) (119.70, 027.41)      /T1_0 i	<|special_separator|>
-(121.79, 020.40) (126.85, 020.40) (126.85, 027.41) (121.79, 027.41)      /T1_0 g	<|special_separator|>
-(126.77, 020.40) (131.41, 020.40) (131.41, 027.41) (126.77, 027.41)      /T1_0 e	<|special_separator|>
-(131.33, 020.40) (136.20, 020.40) (136.20, 027.41) (131.33, 027.41)      /T1_0 n	<|special_separator|>
-(136.12, 020.40) (140.72, 020.40) (140.72, 027.41) (136.12, 027.41)      /T1_0 c	<|special_separator|>
-(140.64, 020.40) (145.28, 020.40) (145.28, 027.41) (140.64, 027.41)      /T1_0 e	<|special_separator|>
-(039.69, 757.67) (046.72, 757.67) (046.72, 766.42) (039.69, 766.42)      /T1_0 A	<|special_separator|>
-(046.62, 757.67) (050.59, 757.67) (050.59, 766.42) (046.62, 766.42)      /T1_0 r	<|special_separator|>
-(050.54, 757.67) (054.23, 757.67) (054.23, 766.42) (050.54, 766.42)      /T1_0 t	<|special_separator|>
-(054.13, 757.67) (056.84, 757.67) (056.84, 766.42) (054.13, 766.42)      /T1_0 i	<|special_separator|>
-(056.74, 757.67) (062.49, 757.67) (062.49, 766.42) (056.74, 766.42)      /T1_0 c	<|special_separator|>
-(062.39, 757.67) (065.52, 757.67) (065.52, 766.42) (062.39, 766.42)      /T1_0 l	<|special_separator|>
-(065.42, 757.67) (071.22, 757.67) (071.22, 766.42) (065.42, 766.42)      /T1_0 e	<|special_separator|>
-(391.20, 757.78) (396.23, 757.78) (396.23, 765.05) (391.20, 765.05)      /T1_1 h	<|special_separator|>
-(396.23, 757.78) (399.60, 757.78) (399.60, 765.05) (396.23, 765.05)      /T1_1 t	<|special_separator|>
-(399.60, 757.78) (402.97, 757.78) (402.97, 765.05) (399.60, 765.05)      /T1_1 t	<|special_separator|>
-(402.97, 757.78) (408.09, 757.78) (408.09, 765.05) (402.97, 765.05)      /T1_1 p	<|special_separator|>
-(408.09, 757.78) (411.77, 757.78) (411.77, 765.05) (408.09, 765.05)      /T1_1 s	<|special_separator|>
-(411.77, 757.78) (414.20, 757.78) (414.20, 765.05) (411.77, 765.05)      /T1_1 :	<|special_separator|>
-(414.20, 757.78) (417.33, 757.78) (417.33, 765.05) (414.20, 765.05)      /T1_1 /	<|special_separator|>
-(416.57, 757.78) (419.69, 757.78) (419.69, 765.05) (416.57, 765.05)      /T1_1 /	<|special_separator|>
-(419.21, 757.78) (424.29, 757.78) (424.29, 765.05) (419.21, 765.05)      /T1_1 d	<|special_separator|>
-(424.29, 757.78) (429.25, 757.78) (429.25, 765.05) (424.29, 765.05)      /T1_1 o	<|special_separator|>
-(429.25, 757.78) (431.73, 757.78) (431.73, 765.05) (429.25, 765.05)      /T1_1 i	<|special_separator|>
-(431.73, 757.78) (434.15, 757.78) (434.15, 765.05) (431.73, 765.05)      /T1_1 .	<|special_separator|>
-(433.95, 757.78) (438.90, 757.78) (438.90, 765.05) (433.95, 765.05)      /T1_1 o	<|special_separator|>
-(438.90, 757.78) (442.49, 757.78) (442.49, 765.05) (438.90, 765.05)      /T1_1 r	<|special_separator|>
-(442.41, 757.78) (447.19, 757.78) (447.19, 765.05) (442.41, 765.05)      /T1_1 g	<|special_separator|>
-(447.23, 757.78) (450.36, 757.78) (450.36, 765.05) (447.23, 765.05)      /T1_1 /	<|special_separator|>
-(450.32, 757.78) (453.95, 757.78) (453.95, 765.05) (450.32, 765.05)      /T1_1 1	<|special_separator|>
-(453.95, 757.78) (459.37, 757.78) (459.37, 765.05) (453.95, 765.05)      /T1_1 0	<|special_separator|>
-(458.97, 757.78) (461.40, 757.78) (461.40, 765.05) (458.97, 765.05)      /T1_1 .	<|special_separator|>
-(460.80, 757.78) (464.43, 757.78) (464.43, 765.05) (460.80, 765.05)      /T1_1 1	<|special_separator|>
-(464.43, 757.78) (469.85, 757.78) (469.85, 765.05) (464.43, 765.05)      /T1_1 0	<|special_separator|>
-(469.81, 757.78) (474.29, 757.78) (474.29, 765.05) (469.81, 765.05)      /T1_1 3	<|special_separator|>
-(474.29, 757.78) (479.13, 757.78) (479.13, 765.05) (474.29, 765.05)      /T1_1 8	<|special_separator|>
-(478.93, 757.78) (482.06, 757.78) (482.06, 765.05) (478.93, 765.05)      /T1_1 /	<|special_separator|>
-(481.62, 757.78) (485.31, 757.78) (485.31, 765.05) (481.62, 765.05)      /T1_1 s	<|special_separator|>
-(485.31, 757.78) (490.21, 757.78) (490.21, 765.05) (485.31, 765.05)      /T1_1 4	<|special_separator|>
-(490.21, 757.78) (494.73, 757.78) (494.73, 765.05) (490.21, 765.05)      /T1_1 2	<|special_separator|>
-(494.76, 757.78) (499.28, 757.78) (499.28, 765.05) (494.76, 765.05)      /T1_1 2	<|special_separator|>
-(499.28, 757.78) (503.60, 757.78) (503.60, 765.05) (499.28, 765.05)      /T1_1 5	<|special_separator|>
-(503.52, 757.78) (508.44, 757.78) (508.44, 765.05) (503.52, 765.05)      /T1_1 6	<|special_separator|>
-(508.40, 757.78) (510.92, 757.78) (510.92, 765.05) (508.40, 765.05)      /T1_1 -	<|special_separator|>
-(510.92, 757.78) (516.34, 757.78) (516.34, 765.05) (510.92, 765.05)      /T1_1 0	<|special_separator|>
-(516.34, 757.78) (520.85, 757.78) (520.85, 765.05) (516.34, 765.05)      /T1_1 2	<|special_separator|>
-(520.81, 757.78) (525.73, 757.78) (525.73, 765.05) (520.81, 765.05)      /T1_1 6	<|special_separator|>
-(525.69, 757.78) (528.20, 757.78) (528.20, 765.05) (525.69, 765.05)      /T1_1 -	<|special_separator|>
-(528.20, 757.78) (533.63, 757.78) (533.63, 765.05) (528.20, 765.05)      /T1_1 0	<|special_separator|>
-(533.63, 757.78) (537.26, 757.78) (537.26, 765.05) (533.63, 765.05)      /T1_1 1	<|special_separator|>
-(537.26, 757.78) (541.77, 757.78) (541.77, 765.05) (537.26, 765.05)      /T1_1 2	<|special_separator|>
-(541.57, 757.78) (546.48, 757.78) (546.48, 765.05) (541.57, 765.05)      /T1_1 4	<|special_separator|>
-(546.48, 757.78) (550.99, 757.78) (550.99, 765.05) (546.48, 765.05)      /T1_1 2	<|special_separator|>
-(550.91, 757.78) (553.42, 757.78) (553.42, 765.05) (550.91, 765.05)      /T1_1 -	<|special_separator|>
-(553.42, 757.78) (558.27, 757.78) (558.27, 765.05) (553.42, 765.05)      /T1_1 8	<|special_separator|>
-(039.69, 173.85) (043.70, 173.85) (043.70, 180.21) (039.69, 180.21)      /T1_1 E	<|special_separator|>
-(043.68, 173.85) (047.81, 173.85) (047.81, 180.21) (043.68, 180.21)      /T1_1 x	<|special_separator|>
-(047.65, 173.85) (050.60, 173.85) (050.60, 180.21) (047.65, 180.21)      /T1_1 t	<|special_separator|>
-(050.55, 173.85) (054.35, 173.85) (054.35, 180.21) (050.55, 180.21)      /T1_1 e	<|special_separator|>
-(054.34, 173.85) (058.74, 173.85) (058.74, 180.21) (054.34, 180.21)      /T1_1 n	<|special_separator|>
-(058.72, 173.85) (063.17, 173.85) (063.17, 180.21) (058.72, 180.21)      /T1_1 d	<|special_separator|>
-(063.15, 173.85) (066.95, 173.85) (066.95, 180.21) (063.15, 180.21)      /T1_1 e	<|special_separator|>
-(066.94, 173.85) (071.39, 173.85) (071.39, 180.21) (066.94, 180.21)      /T1_1 d	<|special_separator|>
-(071.37, 173.85) (072.54, 173.85) (072.54, 180.21) (071.37, 180.21)      /T1_1  	<|special_separator|>
-(072.35, 173.85) (077.63, 173.85) (077.63, 180.21) (072.35, 180.21)      /T1_1 D	<|special_separator|>
-(077.62, 173.85) (081.49, 173.85) (081.49, 180.21) (077.62, 180.21)      /T1_1 a	<|special_separator|>
-(081.47, 173.85) (084.42, 173.85) (084.42, 180.21) (081.47, 180.21)      /T1_1 t	<|special_separator|>
-(084.44, 173.85) (088.31, 173.85) (088.31, 180.21) (084.44, 180.21)      /T1_1 a	<|special_separator|>
-(088.30, 173.85) (089.47, 173.85) (089.47, 180.21) (088.30, 180.21)      /T1_1  	<|special_separator|>
-(089.28, 173.85) (093.33, 173.85) (093.33, 180.21) (089.28, 180.21)      /T1_1 F	<|special_separator|>
-(093.24, 173.85) (095.41, 173.85) (095.41, 180.21) (093.24, 180.21)      /T1_1 i	<|special_separator|>
-(095.40, 173.85) (099.59, 173.85) (099.59, 180.21) (095.40, 180.21)      /T1_1 g	<|special_separator|>
-(099.57, 173.85) (101.69, 173.85) (101.69, 180.21) (099.57, 180.21)      /T1_1 .	<|special_separator|>
-(101.68, 173.85) (102.85, 173.85) (102.85, 180.21) (101.68, 180.21)      /T1_1  	<|special_separator|>
-(102.66, 173.85) (106.61, 173.85) (106.61, 180.21) (102.66, 180.21)      /T1_1 2	<|special_separator|>
-(106.59, 173.85) (107.76, 173.85) (107.76, 180.21) (106.59, 180.21)      /T1_1  	<|special_separator|>
-(107.57, 173.85) (109.25, 173.85) (109.25, 180.21) (107.57, 180.21)      /T1_1 |	<|special_separator|>
-(109.23, 173.85) (110.40, 173.85) (110.40, 180.21) (109.23, 180.21)      /T1_1  	<|special_separator|>
-(110.21, 173.85) (115.23, 173.85) (115.23, 180.21) (110.21, 180.21)      /T1_1 U	<|special_separator|>
-(115.22, 173.85) (121.62, 173.85) (121.62, 180.21) (115.22, 180.21)      /T1_1 M	<|special_separator|>
-(121.60, 173.85) (126.55, 173.85) (126.55, 180.21) (121.60, 180.21)      /T1_1 A	<|special_separator|>
-(126.54, 173.85) (130.91, 173.85) (130.91, 180.21) (126.54, 180.21)      /T1_1 P	<|special_separator|>
-(130.90, 173.85) (132.07, 173.85) (132.07, 180.21) (130.90, 180.21)      /T1_1  	<|special_separator|>
-(131.88, 173.85) (136.36, 173.85) (136.36, 180.21) (131.88, 180.21)      /T1_1 p	<|special_separator|>
-(136.34, 173.85) (139.48, 173.85) (139.48, 180.21) (136.34, 180.21)      /T1_1 r	<|special_separator|>
-(139.43, 173.85) (143.77, 173.85) (143.77, 180.21) (139.43, 180.21)      /T1_1 o	<|special_separator|>
-(143.75, 173.85) (145.91, 173.85) (145.91, 180.21) (143.75, 180.21)      /T1_1 j	<|special_separator|>
-(145.89, 173.85) (149.69, 173.85) (149.69, 180.21) (145.89, 180.21)      /T1_1 e	<|special_separator|>
-(149.68, 173.85) (153.31, 173.85) (153.31, 180.21) (149.68, 180.21)      /T1_1 c	<|special_separator|>
-(153.37, 173.85) (156.31, 173.85) (156.31, 180.21) (153.37, 180.21)      /T1_1 t	<|special_separator|>
-(156.30, 173.85) (158.47, 173.85) (158.47, 180.21) (156.30, 180.21)      /T1_1 i	<|special_separator|>
-(158.45, 173.85) (162.79, 173.85) (162.79, 180.21) (158.45, 180.21)      /T1_1 o	<|special_separator|>
-(162.77, 173.85) (167.17, 173.85) (167.17, 180.21) (162.77, 180.21)      /T1_1 n	<|special_separator|>
-(167.16, 173.85) (168.32, 173.85) (168.32, 180.21) (167.16, 180.21)      /T1_1  	<|special_separator|>
-(168.14, 173.85) (172.47, 173.85) (172.47, 180.21) (168.14, 180.21)      /T1_1 o	<|special_separator|>
-(172.45, 173.85) (175.02, 173.85) (175.02, 180.21) (172.45, 180.21)      /T1_1 f	<|special_separator|>
-(175.00, 173.85) (176.17, 173.85) (176.17, 180.21) (175.00, 180.21)      /T1_1  	<|special_separator|>
-(175.98, 173.85) (178.93, 173.85) (178.93, 180.21) (175.98, 180.21)      /T1_1 t	<|special_separator|>
-(178.92, 173.85) (183.32, 173.85) (183.32, 180.21) (178.92, 180.21)      /T1_1 h	<|special_separator|>
-(183.30, 173.85) (187.11, 173.85) (187.11, 180.21) (183.30, 180.21)      /T1_1 e	<|special_separator|>
-(187.09, 173.85) (188.26, 173.85) (188.26, 180.21) (187.09, 180.21)      /T1_1  	<|special_separator|>
-(188.07, 173.85) (191.25, 173.85) (191.25, 180.21) (188.07, 180.21)      /T1_1 1	<|special_separator|>
-(191.24, 173.85) (195.98, 173.85) (195.98, 180.21) (191.24, 180.21)      /T1_1 0	<|special_separator|>
-(195.97, 173.85) (200.71, 173.85) (200.71, 180.21) (195.97, 180.21)      /T1_1 0	<|special_separator|>
-(200.70, 173.85) (205.45, 173.85) (205.45, 180.21) (200.70, 180.21)      /T1_1 0	<|special_separator|>
-(205.43, 173.85) (207.63, 173.85) (207.63, 180.21) (205.43, 180.21)      /T1_1 -	<|special_separator|>
-(207.69, 173.85) (212.13, 173.85) (212.13, 180.21) (207.69, 180.21)      /T1_1 d	<|special_separator|>
-(212.12, 173.85) (214.29, 173.85) (214.29, 180.21) (212.12, 180.21)      /T1_1 i	<|special_separator|>
-(214.27, 173.85) (220.81, 173.85) (220.81, 180.21) (214.27, 180.21)      /T1_1 m	<|special_separator|>
-(220.79, 173.85) (224.59, 173.85) (224.59, 180.21) (220.79, 180.21)      /T1_1 e	<|special_separator|>
-(224.58, 173.85) (228.97, 173.85) (228.97, 180.21) (224.58, 180.21)      /T1_1 n	<|special_separator|>
-(228.96, 173.85) (232.19, 173.85) (232.19, 180.21) (228.96, 180.21)      /T1_1 s	<|special_separator|>
-(232.17, 173.85) (234.34, 173.85) (234.34, 180.21) (232.17, 180.21)      /T1_1 i	<|special_separator|>
-(234.33, 173.85) (238.66, 173.85) (238.66, 180.21) (234.33, 180.21)      /T1_1 o	<|special_separator|>
-(238.65, 173.85) (243.04, 173.85) (243.04, 180.21) (238.65, 180.21)      /T1_1 n	<|special_separator|>
-(243.03, 173.85) (246.90, 173.85) (246.90, 180.21) (243.03, 180.21)      /T1_1 a	<|special_separator|>
-(246.89, 173.85) (249.09, 173.85) (249.09, 180.21) (246.89, 180.21)      /T1_1 l	<|special_separator|>
-(249.08, 173.85) (250.25, 173.85) (250.25, 180.21) (249.08, 180.21)      /T1_1  	<|special_separator|>
-(250.06, 173.85) (252.62, 173.85) (252.62, 180.21) (250.06, 180.21)      /T1_1 f	<|special_separator|>
-(252.57, 173.85) (256.37, 173.85) (256.37, 180.21) (252.57, 180.21)      /T1_1 e	<|special_separator|>
-(256.36, 173.85) (260.23, 173.85) (260.23, 180.21) (256.36, 180.21)      /T1_1 a	<|special_separator|>
-(260.22, 173.85) (263.16, 173.85) (263.16, 180.21) (260.22, 180.21)      /T1_1 t	<|special_separator|>
-(263.15, 173.85) (267.55, 173.85) (267.55, 180.21) (263.15, 180.21)      /T1_1 u	<|special_separator|>
-(267.54, 173.85) (270.68, 173.85) (270.68, 180.21) (267.54, 180.21)      /T1_1 r	<|special_separator|>
-(270.63, 173.85) (274.43, 173.85) (274.43, 180.21) (270.63, 180.21)      /T1_1 e	<|special_separator|>
-(274.42, 173.85) (275.59, 173.85) (275.59, 180.21) (274.42, 180.21)      /T1_1  	<|special_separator|>
-(275.40, 173.85) (278.62, 173.85) (278.62, 180.21) (275.40, 180.21)      /T1_1 s	<|special_separator|>
-(278.61, 173.85) (283.09, 173.85) (283.09, 180.21) (278.61, 180.21)      /T1_1 p	<|special_separator|>
-(283.04, 173.85) (286.91, 173.85) (286.91, 180.21) (283.04, 180.21)      /T1_1 a	<|special_separator|>
-(286.90, 173.85) (290.53, 173.85) (290.53, 180.21) (286.90, 180.21)      /T1_1 c	<|special_separator|>
-(290.51, 173.85) (294.31, 173.85) (294.31, 180.21) (290.51, 180.21)      /T1_1 e	<|special_separator|>
-(294.23, 173.85) (296.39, 173.85) (296.39, 180.21) (294.23, 180.21)      /T1_1 ,	<|special_separator|>
-(296.37, 173.85) (297.54, 173.85) (297.54, 180.21) (296.37, 180.21)      /T1_1  	<|special_separator|>
-(297.35, 173.85) (299.91, 173.85) (299.91, 180.21) (297.35, 180.21)      /T1_1 f	<|special_separator|>
-(299.90, 173.85) (302.07, 173.85) (302.07, 180.21) (299.90, 180.21)      /T1_1 i	<|special_separator|>
-(302.06, 173.85) (304.26, 173.85) (304.26, 180.21) (302.06, 180.21)      /T1_1 l	<|special_separator|>
-(304.25, 173.85) (307.19, 173.85) (307.19, 180.21) (304.25, 180.21)      /T1_1 t	<|special_separator|>
-(307.15, 173.85) (310.95, 173.85) (310.95, 180.21) (307.15, 180.21)      /T1_1 e	<|special_separator|>
-(310.93, 173.85) (314.07, 173.85) (314.07, 180.21) (310.93, 180.21)      /T1_1 r	<|special_separator|>
-(314.02, 173.85) (317.83, 173.85) (317.83, 180.21) (314.02, 180.21)      /T1_1 e	<|special_separator|>
-(317.81, 173.85) (322.26, 173.85) (322.26, 180.21) (317.81, 180.21)      /T1_1 d	<|special_separator|>
-(322.24, 173.85) (323.41, 173.85) (323.41, 180.21) (322.24, 180.21)      /T1_1  	<|special_separator|>
-(323.22, 173.85) (327.56, 173.85) (327.56, 180.21) (323.22, 180.21)      /T1_1 o	<|special_separator|>
-(327.54, 173.85) (331.94, 173.85) (331.94, 180.21) (327.54, 180.21)      /T1_1 n	<|special_separator|>
-(331.92, 173.85) (333.09, 173.85) (333.09, 180.21) (331.92, 180.21)      /T1_1  	<|special_separator|>
-(332.90, 173.85) (336.53, 173.85) (336.53, 180.21) (332.90, 180.21)      /T1_1 c	<|special_separator|>
-(336.48, 173.85) (340.82, 173.85) (340.82, 180.21) (336.48, 180.21)      /T1_1 o	<|special_separator|>
-(340.80, 173.85) (345.20, 173.85) (345.20, 180.21) (340.80, 180.21)      /T1_1 n	<|special_separator|>
-(345.18, 173.85) (348.13, 173.85) (348.13, 180.21) (345.18, 180.21)      /T1_1 t	<|special_separator|>
-(348.12, 173.85) (351.25, 173.85) (351.25, 180.21) (348.12, 180.21)      /T1_1 r	<|special_separator|>
-(351.21, 173.85) (355.54, 173.85) (355.54, 180.21) (351.21, 180.21)      /T1_1 o	<|special_separator|>
-(355.52, 173.85) (357.73, 173.85) (357.73, 180.21) (355.52, 180.21)      /T1_1 l	<|special_separator|>
-(357.72, 173.85) (358.88, 173.85) (358.88, 180.21) (357.72, 180.21)      /T1_1  	<|special_separator|>
-(358.70, 173.85) (362.32, 173.85) (362.32, 180.21) (358.70, 180.21)      /T1_1 c	<|special_separator|>
-(362.31, 173.85) (366.11, 173.85) (366.11, 180.21) (362.31, 180.21)      /T1_1 e	<|special_separator|>
-(366.06, 173.85) (368.27, 173.85) (368.27, 180.21) (366.06, 180.21)      /T1_1 l	<|special_separator|>
-(368.25, 173.85) (370.46, 173.85) (370.46, 180.21) (368.25, 180.21)      /T1_1 l	<|special_separator|>
-(370.44, 173.85) (373.67, 173.85) (373.67, 180.21) (370.44, 180.21)      /T1_1 s	<|special_separator|>
-(373.66, 173.85) (374.83, 173.85) (374.83, 180.21) (373.66, 180.21)      /T1_1  	<|special_separator|>
-(374.64, 173.85) (378.51, 173.85) (378.51, 180.21) (374.64, 180.21)      /T1_1 a	<|special_separator|>
-(378.49, 173.85) (382.89, 173.85) (382.89, 180.21) (378.49, 180.21)      /T1_1 n	<|special_separator|>
-(382.88, 173.85) (387.32, 173.85) (387.32, 180.21) (382.88, 180.21)      /T1_1 d	<|special_separator|>
-(387.31, 173.85) (388.48, 173.85) (388.48, 180.21) (387.31, 180.21)      /T1_1  	<|special_separator|>
-(388.29, 173.85) (391.91, 173.85) (391.91, 180.21) (388.29, 180.21)      /T1_1 c	<|special_separator|>
-(391.90, 173.85) (395.70, 173.85) (395.70, 180.21) (391.90, 180.21)      /T1_1 e	<|special_separator|>
-(395.65, 173.85) (397.86, 173.85) (397.86, 180.21) (395.65, 180.21)      /T1_1 l	<|special_separator|>
-(397.84, 173.85) (400.05, 173.85) (400.05, 180.21) (397.84, 180.21)      /T1_1 l	<|special_separator|>
-(400.04, 173.85) (403.26, 173.85) (403.26, 180.21) (400.04, 180.21)      /T1_1 s	<|special_separator|>
-(403.25, 173.85) (404.42, 173.85) (404.42, 180.21) (403.25, 180.21)      /T1_1  	<|special_separator|>
-(404.23, 173.85) (407.18, 173.85) (407.18, 180.21) (404.23, 180.21)      /T1_1 t	<|special_separator|>
-(407.16, 173.85) (410.30, 173.85) (410.30, 180.21) (407.16, 180.21)      /T1_1 r	<|special_separator|>
-(410.25, 173.85) (414.05, 173.85) (414.05, 180.21) (410.25, 180.21)      /T1_1 e	<|special_separator|>
-(414.04, 173.85) (417.91, 173.85) (417.91, 180.21) (414.04, 180.21)      /T1_1 a	<|special_separator|>
-(417.90, 173.85) (420.84, 173.85) (420.84, 180.21) (417.90, 180.21)      /T1_1 t	<|special_separator|>
-(420.80, 173.85) (424.60, 173.85) (424.60, 180.21) (420.80, 180.21)      /T1_1 e	<|special_separator|>
-(424.58, 173.85) (429.03, 173.85) (429.03, 180.21) (424.58, 180.21)      /T1_1 d	<|special_separator|>
-(429.01, 173.85) (430.18, 173.85) (430.18, 180.21) (429.01, 180.21)      /T1_1  	<|special_separator|>
-(429.99, 173.85) (435.49, 173.85) (435.49, 180.21) (429.99, 180.21)      /T1_1 w	<|special_separator|>
-(435.44, 173.85) (437.61, 173.85) (437.61, 180.21) (435.44, 180.21)      /T1_1 i	<|special_separator|>
-(437.60, 173.85) (440.54, 173.85) (440.54, 180.21) (437.60, 180.21)      /T1_1 t	<|special_separator|>
-(440.53, 173.85) (444.93, 173.85) (444.93, 180.21) (440.53, 180.21)      /T1_1 h	<|special_separator|>
-(444.92, 173.85) (446.09, 173.85) (446.09, 180.21) (444.92, 180.21)      /T1_1  	<|special_separator|>
-(445.90, 173.85) (450.34, 173.85) (450.34, 180.21) (445.90, 180.21)      /T1_1 d	<|special_separator|>
-(450.33, 173.85) (452.50, 173.85) (452.50, 180.21) (450.33, 180.21)      /T1_1 i	<|special_separator|>
-(452.49, 173.85) (455.05, 173.85) (455.05, 180.21) (452.49, 180.21)      /T1_1 f	<|special_separator|>
-(455.07, 173.85) (457.63, 173.85) (457.63, 180.21) (455.07, 180.21)      /T1_1 f	<|special_separator|>
-(457.59, 173.85) (461.39, 173.85) (461.39, 180.21) (457.59, 180.21)      /T1_1 e	<|special_separator|>
-(461.37, 173.85) (464.51, 173.85) (464.51, 180.21) (461.37, 180.21)      /T1_1 r	<|special_separator|>
-(464.46, 173.85) (468.26, 173.85) (468.26, 180.21) (464.46, 180.21)      /T1_1 e	<|special_separator|>
-(468.25, 173.85) (472.65, 173.85) (472.65, 180.21) (468.25, 180.21)      /T1_1 n	<|special_separator|>
-(472.63, 173.85) (475.58, 173.85) (475.58, 180.21) (472.63, 180.21)      /T1_1 t	<|special_separator|>
-(475.57, 173.85) (476.73, 173.85) (476.73, 180.21) (475.57, 180.21)      /T1_1  	<|special_separator|>
-(476.55, 173.85) (480.99, 173.85) (480.99, 180.21) (476.55, 180.21)      /T1_1 d	<|special_separator|>
-(480.98, 173.85) (485.31, 173.85) (485.31, 180.21) (480.98, 180.21)      /T1_1 o	<|special_separator|>
-(485.26, 173.85) (488.49, 173.85) (488.49, 180.21) (485.26, 180.21)      /T1_1 s	<|special_separator|>
-(488.48, 173.85) (492.35, 173.85) (492.35, 180.21) (488.48, 180.21)      /T1_1 a	<|special_separator|>
-(492.33, 173.85) (496.52, 173.85) (496.52, 180.21) (492.33, 180.21)      /T1_1 g	<|special_separator|>
-(496.40, 173.85) (500.20, 173.85) (500.20, 180.21) (496.40, 180.21)      /T1_1 e	<|special_separator|>
-(500.19, 173.85) (503.42, 173.85) (503.42, 180.21) (500.19, 180.21)      /T1_1 s	<|special_separator|>
-(503.40, 173.85) (505.52, 173.85) (505.52, 180.21) (503.40, 180.21)      /T1_1 .	<|special_separator|>
-(505.51, 173.72) (506.74, 173.72) (506.74, 180.21) (505.51, 180.21)      /T1_2  	<|special_separator|>
-(506.55, 173.72) (513.10, 173.72) (513.10, 180.21) (506.55, 180.21)      /T1_2 W	<|special_separator|>
-(512.73, 173.72) (516.46, 173.72) (516.46, 180.21) (512.73, 180.21)      /T1_2 e	<|special_separator|>
-(516.44, 173.72) (517.68, 173.72) (517.68, 180.21) (516.44, 180.21)      /T1_2  	<|special_separator|>
-(517.48, 173.72) (521.06, 173.72) (521.06, 180.21) (517.48, 180.21)      /T1_2 c	<|special_separator|>
-(521.05, 173.72) (524.68, 173.72) (524.68, 180.21) (521.05, 180.21)      /T1_2 a	<|special_separator|>
-(524.66, 173.72) (528.84, 173.72) (528.84, 180.21) (524.66, 180.21)      /T1_2 n	<|special_separator|>
-(528.83, 173.72) (530.06, 173.72) (530.06, 180.21) (528.83, 180.21)      /T1_2  	<|special_separator|>
-(529.86, 173.72) (534.10, 173.72) (534.10, 180.21) (529.86, 180.21)      /T1_2 o	<|special_separator|>
-(534.08, 173.72) (538.33, 173.72) (538.33, 180.21) (534.08, 180.21)      /T1_2 b	<|special_separator|>
-(538.32, 173.72) (541.49, 173.72) (541.49, 180.21) (538.32, 180.21)      /T1_2 s	<|special_separator|>
-(541.41, 173.72) (545.13, 173.72) (545.13, 180.21) (541.41, 180.21)      /T1_2 e	<|special_separator|>
-(545.12, 173.72) (548.02, 173.72) (548.02, 180.21) (545.12, 180.21)      /T1_2 r	<|special_separator|>
-(548.15, 173.72) (551.87, 173.72) (551.87, 180.21) (548.15, 180.21)      /T1_2 v	<|special_separator|>
-(551.72, 173.72) (555.44, 173.72) (555.44, 180.21) (551.72, 180.21)      /T1_2 e	<|special_separator|>
-(555.43, 173.72) (556.66, 173.72) (556.66, 180.21) (555.43, 180.21)      /T1_2  	<|special_separator|>
-(556.46, 173.72) (560.09, 173.72) (560.09, 180.21) (556.46, 180.21)      /T1_2 a	<|special_separator|>
-(560.09, 173.72) (561.32, 173.72) (561.32, 180.21) (560.09, 180.21)      /T1_2  	<|special_separator|>
-(039.69, 163.71) (043.76, 163.71) (043.76, 170.21) (039.69, 170.21)      /T1_2 g	<|special_separator|>
-(043.75, 163.71) (046.65, 163.71) (046.65, 170.21) (043.75, 170.21)      /T1_2 r	<|special_separator|>
-(046.53, 163.71) (050.26, 163.71) (050.26, 170.21) (046.53, 170.21)      /T1_2 e	<|special_separator|>
-(050.24, 163.71) (053.87, 163.71) (053.87, 170.21) (050.24, 170.21)      /T1_2 a	<|special_separator|>
-(053.86, 163.71) (056.54, 163.71) (056.54, 170.21) (053.86, 170.21)      /T1_2 t	<|special_separator|>
-(056.45, 163.71) (060.18, 163.71) (060.18, 170.21) (056.45, 170.21)      /T1_2 e	<|special_separator|>
-(060.16, 163.71) (063.07, 163.71) (063.07, 170.21) (060.16, 170.21)      /T1_2 r	<|special_separator|>
-(063.06, 163.71) (064.29, 163.71) (064.29, 170.21) (063.06, 170.21)      /T1_2  	<|special_separator|>
-(064.09, 163.71) (068.42, 163.71) (068.42, 170.21) (064.09, 170.21)      /T1_2 p	<|special_separator|>
-(068.40, 163.71) (072.13, 163.71) (072.13, 170.21) (068.40, 170.21)      /T1_2 e	<|special_separator|>
-(072.11, 163.71) (075.02, 163.71) (075.02, 170.21) (072.11, 170.21)      /T1_2 r	<|special_separator|>
-(075.08, 163.71) (077.76, 163.71) (077.76, 170.21) (075.08, 170.21)      /T1_2 t	<|special_separator|>
-(077.74, 163.71) (081.94, 163.71) (081.94, 170.21) (077.74, 170.21)      /T1_2 u	<|special_separator|>
-(081.92, 163.71) (084.83, 163.71) (084.83, 170.21) (081.92, 170.21)      /T1_2 r	<|special_separator|>
-(084.81, 163.71) (089.06, 163.71) (089.06, 170.21) (084.81, 170.21)      /T1_2 b	<|special_separator|>
-(088.98, 163.71) (092.61, 163.71) (092.61, 170.21) (088.98, 170.21)      /T1_2 a	<|special_separator|>
-(092.59, 163.71) (095.27, 163.71) (095.27, 170.21) (092.59, 170.21)      /T1_2 t	<|special_separator|>
-(095.26, 163.71) (097.26, 163.71) (097.26, 170.21) (095.26, 170.21)      /T1_2 i	<|special_separator|>
-(097.25, 163.71) (101.48, 163.71) (101.48, 170.21) (097.25, 170.21)      /T1_2 o	<|special_separator|>
-(101.47, 163.71) (105.65, 163.71) (105.65, 170.21) (101.47, 170.21)      /T1_2 n	<|special_separator|>
-(105.63, 163.71) (106.86, 163.71) (106.86, 170.21) (105.63, 170.21)      /T1_2  	<|special_separator|>
-(106.67, 163.71) (110.39, 163.71) (110.39, 170.21) (106.67, 170.21)      /T1_2 e	<|special_separator|>
-(110.35, 163.71) (112.87, 163.71) (112.87, 170.21) (110.35, 170.21)      /T1_2 f	<|special_separator|>
-(112.83, 163.71) (115.35, 163.71) (115.35, 170.21) (112.83, 170.21)      /T1_2 f	<|special_separator|>
-(115.20, 163.71) (118.92, 163.71) (118.92, 170.21) (115.20, 170.21)      /T1_2 e	<|special_separator|>
-(118.91, 163.71) (122.49, 163.71) (122.49, 170.21) (118.91, 170.21)      /T1_2 c	<|special_separator|>
-(122.52, 163.71) (125.20, 163.71) (125.20, 170.21) (122.52, 170.21)      /T1_2 t	<|special_separator|>
-(125.18, 163.71) (126.42, 163.71) (126.42, 170.21) (125.18, 170.21)      /T1_2  	<|special_separator|>
-(126.22, 163.71) (131.48, 163.71) (131.48, 170.21) (126.22, 170.21)      /T1_2 w	<|special_separator|>
-(131.47, 163.71) (133.47, 163.71) (133.47, 170.21) (131.47, 170.21)      /T1_2 i	<|special_separator|>
-(133.46, 163.71) (136.14, 163.71) (136.14, 170.21) (133.46, 170.21)      /T1_2 t	<|special_separator|>
-(136.12, 163.71) (140.31, 163.71) (140.31, 170.21) (136.12, 170.21)      /T1_2 h	<|special_separator|>
-(140.30, 163.71) (141.53, 163.71) (141.53, 170.21) (140.30, 170.21)      /T1_2  	<|special_separator|>
-(141.33, 163.71) (145.52, 163.71) (145.52, 170.21) (141.33, 170.21)      /T1_2 h	<|special_separator|>
-(145.50, 163.71) (147.51, 163.71) (147.51, 170.21) (145.50, 170.21)      /T1_2 i	<|special_separator|>
-(147.49, 163.71) (151.57, 163.71) (151.57, 170.21) (147.49, 170.21)      /T1_2 g	<|special_separator|>
-(151.55, 163.71) (155.74, 163.71) (155.74, 170.21) (151.55, 170.21)      /T1_2 h	<|special_separator|>
-(155.72, 163.71) (159.45, 163.71) (159.45, 170.21) (155.72, 170.21)      /T1_2 e	<|special_separator|>
-(159.43, 163.71) (162.34, 163.71) (162.34, 170.21) (159.43, 170.21)      /T1_2 r	<|special_separator|>
-(162.33, 163.71) (163.56, 163.71) (163.56, 170.21) (162.33, 170.21)      /T1_2  	<|special_separator|>
-(163.36, 163.71) (167.67, 163.71) (167.67, 170.21) (163.36, 170.21)      /T1_2 d	<|special_separator|>
-(167.65, 163.71) (171.89, 163.71) (171.89, 170.21) (167.65, 170.21)      /T1_2 o	<|special_separator|>
-(171.84, 163.71) (175.01, 163.71) (175.01, 170.21) (171.84, 170.21)      /T1_2 s	<|special_separator|>
-(175.00, 163.71) (178.62, 163.71) (178.62, 170.21) (175.00, 170.21)      /T1_2 a	<|special_separator|>
-(178.61, 163.71) (182.68, 163.71) (182.68, 170.21) (178.61, 170.21)      /T1_2 g	<|special_separator|>
-(182.53, 163.71) (186.25, 163.71) (186.25, 170.21) (182.53, 170.21)      /T1_2 e	<|special_separator|>
-(186.24, 163.71) (188.13, 163.71) (188.13, 170.21) (186.24, 170.21)      /T1_2 .	<|special_separator|>
-(188.12, 163.71) (189.35, 163.71) (189.35, 170.21) (188.12, 170.21)      /T1_2  	<|special_separator|>
-(189.15, 163.71) (195.45, 163.71) (195.45, 170.21) (189.15, 170.21)      /T1_2 M	<|special_separator|>
-(195.44, 163.71) (199.67, 163.71) (199.67, 170.21) (195.44, 170.21)      /T1_2 o	<|special_separator|>
-(199.66, 163.71) (202.56, 163.71) (202.56, 170.21) (199.66, 170.21)      /T1_2 r	<|special_separator|>
-(202.45, 163.71) (206.17, 163.71) (206.17, 170.21) (202.45, 170.21)      /T1_2 e	<|special_separator|>
-(206.16, 163.71) (210.39, 163.71) (210.39, 170.21) (206.16, 170.21)      /T1_2 o	<|special_separator|>
-(210.24, 163.71) (213.96, 163.71) (213.96, 170.21) (210.24, 170.21)      /T1_2 v	<|special_separator|>
-(213.80, 163.71) (217.53, 163.71) (217.53, 170.21) (213.80, 170.21)      /T1_2 e	<|special_separator|>
-(217.51, 163.71) (220.42, 163.71) (220.42, 170.21) (217.51, 170.21)      /T1_2 r	<|special_separator|>
-(220.16, 163.71) (222.09, 163.71) (222.09, 170.21) (220.16, 170.21)      /T1_2 ,	<|special_separator|>
-(222.07, 163.71) (223.30, 163.71) (223.30, 170.21) (222.07, 170.21)      /T1_2  	<|special_separator|>
-(223.11, 163.71) (225.79, 163.71) (225.79, 170.21) (223.11, 170.21)      /T1_2 t	<|special_separator|>
-(225.77, 163.71) (229.96, 163.71) (229.96, 170.21) (225.77, 170.21)      /T1_2 h	<|special_separator|>
-(229.95, 163.71) (233.67, 163.71) (233.67, 170.21) (229.95, 170.21)      /T1_2 e	<|special_separator|>
-(233.66, 163.71) (234.89, 163.71) (234.89, 170.21) (233.66, 170.21)      /T1_2  	<|special_separator|>
-(234.69, 163.71) (237.37, 163.71) (237.37, 170.21) (234.69, 170.21)      /T1_2 t	<|special_separator|>
-(237.36, 163.71) (241.55, 163.71) (241.55, 170.21) (237.36, 170.21)      /T1_2 h	<|special_separator|>
-(241.53, 163.71) (244.44, 163.71) (244.44, 170.21) (241.53, 170.21)      /T1_2 r	<|special_separator|>
-(244.32, 163.71) (248.04, 163.71) (248.04, 170.21) (244.32, 170.21)      /T1_2 e	<|special_separator|>
-(248.03, 163.71) (251.75, 163.71) (251.75, 170.21) (248.03, 170.21)      /T1_2 e	<|special_separator|>
-(251.74, 163.71) (252.97, 163.71) (252.97, 170.21) (251.74, 170.21)      /T1_2  	<|special_separator|>
-(252.77, 163.71) (256.36, 163.71) (256.36, 170.21) (252.77, 170.21)      /T1_2 c	<|special_separator|>
-(256.34, 163.71) (258.34, 163.71) (258.34, 170.21) (256.34, 170.21)      /T1_2 l	<|special_separator|>
-(258.33, 163.71) (262.52, 163.71) (262.52, 170.21) (258.33, 170.21)      /T1_2 u	<|special_separator|>
-(262.50, 163.71) (265.68, 163.71) (265.68, 170.21) (262.50, 170.21)      /T1_2 s	<|special_separator|>
-(265.63, 163.71) (268.31, 163.71) (268.31, 170.21) (265.63, 170.21)      /T1_2 t	<|special_separator|>
-(268.23, 163.71) (271.95, 163.71) (271.95, 170.21) (268.23, 170.21)      /T1_2 e	<|special_separator|>
-(271.94, 163.71) (274.84, 163.71) (274.84, 170.21) (271.94, 170.21)      /T1_2 r	<|special_separator|>
-(274.83, 163.71) (278.00, 163.71) (278.00, 170.21) (274.83, 170.21)      /T1_2 s	<|special_separator|>
-(277.99, 163.71) (279.22, 163.71) (279.22, 170.21) (277.99, 170.21)      /T1_2  	<|special_separator|>
-(279.02, 163.71) (282.65, 163.71) (282.65, 170.21) (279.02, 170.21)      /T1_2 a	<|special_separator|>
-(282.63, 163.71) (285.54, 163.71) (285.54, 170.21) (282.63, 170.21)      /T1_2 r	<|special_separator|>
-(285.42, 163.71) (289.14, 163.71) (289.14, 170.21) (285.42, 170.21)      /T1_2 e	<|special_separator|>
-(289.13, 163.71) (290.36, 163.71) (290.36, 170.21) (289.13, 170.21)      /T1_2  	<|special_separator|>
-(290.17, 163.71) (293.79, 163.71) (293.79, 170.21) (290.17, 170.21)      /T1_2 a	<|special_separator|>
-(293.78, 163.71) (296.95, 163.71) (296.95, 170.21) (293.78, 170.21)      /T1_2 s	<|special_separator|>
-(296.87, 163.71) (300.04, 163.71) (300.04, 170.21) (296.87, 170.21)      /T1_2 s	<|special_separator|>
-(300.02, 163.71) (304.26, 163.71) (304.26, 170.21) (300.02, 170.21)      /T1_2 o	<|special_separator|>
-(304.25, 163.71) (307.83, 163.71) (307.83, 170.21) (304.25, 170.21)      /T1_2 c	<|special_separator|>
-(307.82, 163.71) (309.82, 163.71) (309.82, 170.21) (307.82, 170.21)      /T1_2 i	<|special_separator|>
-(309.80, 163.71) (313.43, 163.71) (313.43, 170.21) (309.80, 170.21)      /T1_2 a	<|special_separator|>
-(313.42, 163.71) (316.10, 163.71) (316.10, 170.21) (313.42, 170.21)      /T1_2 t	<|special_separator|>
-(316.01, 163.71) (319.74, 163.71) (319.74, 170.21) (316.01, 170.21)      /T1_2 e	<|special_separator|>
-(319.72, 163.71) (324.03, 163.71) (324.03, 170.21) (319.72, 170.21)      /T1_2 d	<|special_separator|>
-(324.02, 163.71) (325.25, 163.71) (325.25, 170.21) (324.02, 170.21)      /T1_2  	<|special_separator|>
-(325.05, 163.71) (330.32, 163.71) (330.32, 170.21) (325.05, 170.21)      /T1_2 w	<|special_separator|>
-(330.30, 163.71) (332.30, 163.71) (332.30, 170.21) (330.30, 170.21)      /T1_2 i	<|special_separator|>
-(332.29, 163.71) (334.97, 163.71) (334.97, 170.21) (332.29, 170.21)      /T1_2 t	<|special_separator|>
-(334.96, 163.71) (339.14, 163.71) (339.14, 170.21) (334.96, 170.21)      /T1_2 h	<|special_separator|>
-(339.13, 163.71) (340.36, 163.71) (340.36, 170.21) (339.13, 170.21)      /T1_2  	<|special_separator|>
-(340.16, 163.71) (342.85, 163.71) (342.85, 170.21) (340.16, 170.21)      /T1_2 t	<|special_separator|>
-(342.83, 163.71) (347.02, 163.71) (347.02, 170.21) (342.83, 170.21)      /T1_2 h	<|special_separator|>
-(347.00, 163.71) (350.73, 163.71) (350.73, 170.21) (347.00, 170.21)      /T1_2 e	<|special_separator|>
-(350.71, 163.71) (351.95, 163.71) (351.95, 170.21) (350.71, 170.21)      /T1_2  	<|special_separator|>
-(351.75, 163.71) (354.43, 163.71) (354.43, 170.21) (351.75, 170.21)      /T1_2 t	<|special_separator|>
-(354.42, 163.71) (358.60, 163.71) (358.60, 170.21) (354.42, 170.21)      /T1_2 h	<|special_separator|>
-(358.59, 163.71) (361.49, 163.71) (361.49, 170.21) (358.59, 170.21)      /T1_2 r	<|special_separator|>
-(361.38, 163.71) (365.10, 163.71) (365.10, 170.21) (361.38, 170.21)      /T1_2 e	<|special_separator|>
-(365.09, 163.71) (368.81, 163.71) (368.81, 170.21) (365.09, 170.21)      /T1_2 e	<|special_separator|>
-(368.80, 163.71) (370.03, 163.71) (370.03, 170.21) (368.80, 170.21)      /T1_2  	<|special_separator|>
-(369.83, 163.71) (374.14, 163.71) (374.14, 170.21) (369.83, 170.21)      /T1_2 d	<|special_separator|>
-(374.12, 163.71) (376.13, 163.71) (376.13, 170.21) (374.12, 170.21)      /T1_2 i	<|special_separator|>
-(376.11, 163.71) (378.64, 163.71) (378.64, 170.21) (376.11, 170.21)      /T1_2 f	<|special_separator|>
-(378.59, 163.71) (381.12, 163.71) (381.12, 170.21) (378.59, 170.21)      /T1_2 f	<|special_separator|>
-(380.97, 163.71) (384.69, 163.71) (384.69, 170.21) (380.97, 170.21)      /T1_2 e	<|special_separator|>
-(384.68, 163.71) (387.58, 163.71) (387.58, 170.21) (384.68, 170.21)      /T1_2 r	<|special_separator|>
-(387.46, 163.71) (391.19, 163.71) (391.19, 170.21) (387.46, 170.21)      /T1_2 e	<|special_separator|>
-(391.17, 163.71) (395.35, 163.71) (395.35, 170.21) (391.17, 170.21)      /T1_2 n	<|special_separator|>
-(395.34, 163.71) (398.02, 163.71) (398.02, 170.21) (395.34, 170.21)      /T1_2 t	<|special_separator|>
-(398.01, 163.71) (399.24, 163.71) (399.24, 170.21) (398.01, 170.21)      /T1_2  	<|special_separator|>
-(399.04, 163.71) (402.63, 163.71) (402.63, 170.21) (399.04, 170.21)      /T1_2 c	<|special_separator|>
-(402.58, 163.71) (406.30, 163.71) (406.30, 170.21) (402.58, 170.21)      /T1_2 e	<|special_separator|>
-(406.29, 163.71) (408.28, 163.71) (408.28, 170.21) (406.29, 170.21)      /T1_2 l	<|special_separator|>
-(408.27, 163.71) (410.26, 163.71) (410.26, 170.21) (408.27, 170.21)      /T1_2 l	<|special_separator|>
-(410.25, 163.71) (411.48, 163.71) (411.48, 170.21) (410.25, 170.21)      /T1_2  	<|special_separator|>
-(411.29, 163.71) (413.97, 163.71) (413.97, 170.21) (411.29, 170.21)      /T1_2 t	<|special_separator|>
-(413.95, 163.71) (417.64, 163.71) (417.64, 170.21) (413.95, 170.21)      /T1_2 y	<|special_separator|>
-(417.62, 163.71) (421.95, 163.71) (421.95, 170.21) (417.62, 170.21)      /T1_2 p	<|special_separator|>
-(421.93, 163.71) (425.66, 163.71) (425.66, 170.21) (421.93, 170.21)      /T1_2 e	<|special_separator|>
-(425.64, 163.71) (428.81, 163.71) (428.81, 170.21) (425.64, 170.21)      /T1_2 s	<|special_separator|>
-(428.80, 163.71) (430.03, 163.71) (430.03, 170.21) (428.80, 170.21)      /T1_2  	<|special_separator|>
-(429.84, 163.71) (432.41, 163.71) (432.41, 170.21) (429.84, 170.21)      /T1_2 (	<|special_separator|>
-(432.40, 163.71) (435.57, 163.71) (435.57, 170.21) (432.40, 170.21)      /T1_2 s	<|special_separator|>
-(435.49, 163.71) (439.21, 163.71) (439.21, 170.21) (435.49, 170.21)      /T1_2 e	<|special_separator|>
-(439.20, 163.71) (442.92, 163.71) (442.92, 170.21) (439.20, 170.21)      /T1_2 e	<|special_separator|>
-(442.91, 163.71) (444.14, 163.71) (444.14, 170.21) (442.91, 170.21)      /T1_2  	<|special_separator|>
-(443.94, 163.71) (448.19, 163.71) (448.19, 170.21) (443.94, 170.21)      /T1_2 b	<|special_separator|>
-(448.18, 163.71) (450.17, 163.71) (450.17, 170.21) (448.18, 170.21)      /T1_2 l	<|special_separator|>
-(450.16, 163.71) (453.78, 163.71) (453.78, 170.21) (450.16, 170.21)      /T1_2 a	<|special_separator|>
-(453.77, 163.71) (457.35, 163.71) (457.35, 170.21) (453.77, 170.21)      /T1_2 c	<|special_separator|>
-(457.34, 163.71) (461.16, 163.71) (461.16, 170.21) (457.34, 170.21)      /T1_2 k	<|special_separator|>
-(461.14, 163.71) (462.37, 163.71) (462.37, 170.21) (461.14, 170.21)      /T1_2  	<|special_separator|>
-(462.18, 163.71) (464.86, 163.71) (464.86, 170.21) (462.18, 170.21)      /T1_2 t	<|special_separator|>
-(464.78, 163.71) (468.50, 163.71) (468.50, 170.21) (464.78, 170.21)      /T1_2 e	<|special_separator|>
-(468.38, 163.71) (472.24, 163.71) (472.24, 170.21) (468.38, 170.21)      /T1_2 x	<|special_separator|>
-(472.16, 163.71) (474.84, 163.71) (474.84, 170.21) (472.16, 170.21)      /T1_2 t	<|special_separator|>
-(474.83, 163.71) (477.40, 163.71) (477.40, 170.21) (474.83, 170.21)      /T1_2 )	<|special_separator|>
-(477.39, 163.71) (479.28, 163.71) (479.28, 170.21) (477.39, 170.21)      /T1_2 .

groundtruth/parser/font_11.pdf.page_no_15.py.json.line.txt

@@ -1,6 +0,0 @@
-(039.69, 020.40) (145.28, 020.40) (145.28, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 173.85) (505.52, 173.85) (505.52, 180.21) (039.69, 180.21)      /T1_1 Extended Data Fig. 2 | UMAP projection of the 1000-dimensional feature space, filtered on control cells and cells treated with different dosages.	<|special_separator|>
-(505.51, 173.72) (561.32, 173.72) (561.32, 180.21) (505.51, 180.21)      /T1_2  We can observe a 	<|special_separator|>
-(039.69, 163.71) (479.28, 163.71) (479.28, 170.21) (039.69, 170.21)      /T1_2 greater perturbation effect with higher dosage. Moreover, the three clusters are associated with the three different cell types (see black text).

groundtruth/parser/font_11.pdf.page_no_15.py.json.word.txt

@@ -1,7 +0,0 @@
-(039.69, 020.40) (065.17, 020.40) (065.17, 027.41) (039.69, 027.41)      /T1_0 Nature	<|special_separator|>
-(066.63, 020.40) (145.28, 020.40) (145.28, 027.41) (066.63, 027.41)      /T1_0 MachineIntelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 173.85) (505.52, 173.85) (505.52, 180.21) (039.69, 180.21)      /T1_1 ExtendedDataFig.2|UMAPprojectionofthe1000-dimensionalfeaturespace,filteredoncontrolcellsandcellstreatedwithdifferentdosages.	<|special_separator|>
-(506.55, 173.72) (560.09, 173.72) (560.09, 180.21) (506.55, 180.21)      /T1_2 Wecanobservea	<|special_separator|>
-(039.69, 163.71) (479.28, 163.71) (479.28, 170.21) (039.69, 170.21)      /T1_2 greaterperturbationeffectwithhigherdosage.Moreover,thethreeclustersareassociatedwiththethreedifferentcelltypes(seeblacktext).

groundtruth/parser/font_11.pdf.page_no_16.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7125765aaa0df4c3b066076a4b3dbd99cd6f38ac7fde0f9d3a1f24f53515bbff
-size 301955

groundtruth/parser/font_11.pdf.page_no_16.py.json.char.txt

@@ -1,459 +0,0 @@
-(039.69, 020.40) (045.65, 020.40) (045.65, 027.41) (039.69, 027.41)      /T1_0 N	<|special_separator|>
-(045.53, 020.40) (049.97, 020.40) (049.97, 027.41) (045.53, 027.41)      /T1_0 a	<|special_separator|>
-(049.89, 020.40) (052.84, 020.40) (052.84, 027.41) (049.89, 027.41)      /T1_0 t	<|special_separator|>
-(052.76, 020.40) (057.60, 020.40) (057.60, 027.41) (052.76, 027.41)      /T1_0 u	<|special_separator|>
-(057.52, 020.40) (060.69, 020.40) (060.69, 027.41) (057.52, 027.41)      /T1_0 r	<|special_separator|>
-(060.53, 020.40) (065.17, 020.40) (065.17, 027.41) (060.53, 027.41)      /T1_0 e	<|special_separator|>
-(065.09, 020.40) (067.00, 020.40) (067.00, 027.41) (065.09, 027.41)      /T1_0  	<|special_separator|>
-(066.63, 020.40) (073.75, 020.40) (073.75, 027.41) (066.63, 027.41)      /T1_0 M	<|special_separator|>
-(073.63, 020.40) (078.07, 020.40) (078.07, 027.41) (073.63, 027.41)      /T1_0 a	<|special_separator|>
-(077.99, 020.40) (082.59, 020.40) (082.59, 027.41) (077.99, 027.41)      /T1_0 c	<|special_separator|>
-(082.51, 020.40) (087.37, 020.40) (087.37, 027.41) (082.51, 027.41)      /T1_0 h	<|special_separator|>
-(087.29, 020.40) (089.46, 020.40) (089.46, 027.41) (087.29, 027.41)      /T1_0 i	<|special_separator|>
-(089.38, 020.40) (094.25, 020.40) (094.25, 027.41) (089.38, 027.41)      /T1_0 n	<|special_separator|>
-(094.17, 020.40) (098.81, 020.40) (098.81, 027.41) (094.17, 027.41)      /T1_0 e	<|special_separator|>
-(098.73, 020.40) (100.63, 020.40) (100.63, 027.41) (098.73, 027.41)      /T1_0  	<|special_separator|>
-(100.26, 020.40) (102.72, 020.40) (102.72, 027.41) (100.26, 027.41)      /T1_0 I	<|special_separator|>
-(102.64, 020.40) (107.50, 020.40) (107.50, 027.41) (102.64, 027.41)      /T1_0 n	<|special_separator|>
-(107.42, 020.40) (110.37, 020.40) (110.37, 027.41) (107.42, 027.41)      /T1_0 t	<|special_separator|>
-(110.29, 020.40) (114.93, 020.40) (114.93, 027.41) (110.29, 027.41)      /T1_0 e	<|special_separator|>
-(114.85, 020.40) (117.36, 020.40) (117.36, 027.41) (114.85, 027.41)      /T1_0 l	<|special_separator|>
-(117.28, 020.40) (119.78, 020.40) (119.78, 027.41) (117.28, 027.41)      /T1_0 l	<|special_separator|>
-(119.70, 020.40) (121.87, 020.40) (121.87, 027.41) (119.70, 027.41)      /T1_0 i	<|special_separator|>
-(121.79, 020.40) (126.85, 020.40) (126.85, 027.41) (121.79, 027.41)      /T1_0 g	<|special_separator|>
-(126.77, 020.40) (131.41, 020.40) (131.41, 027.41) (126.77, 027.41)      /T1_0 e	<|special_separator|>
-(131.33, 020.40) (136.20, 020.40) (136.20, 027.41) (131.33, 027.41)      /T1_0 n	<|special_separator|>
-(136.12, 020.40) (140.72, 020.40) (140.72, 027.41) (136.12, 027.41)      /T1_0 c	<|special_separator|>
-(140.64, 020.40) (145.28, 020.40) (145.28, 027.41) (140.64, 027.41)      /T1_0 e	<|special_separator|>
-(039.69, 757.67) (046.72, 757.67) (046.72, 766.42) (039.69, 766.42)      /T1_0 A	<|special_separator|>
-(046.62, 757.67) (050.59, 757.67) (050.59, 766.42) (046.62, 766.42)      /T1_0 r	<|special_separator|>
-(050.54, 757.67) (054.23, 757.67) (054.23, 766.42) (050.54, 766.42)      /T1_0 t	<|special_separator|>
-(054.13, 757.67) (056.84, 757.67) (056.84, 766.42) (054.13, 766.42)      /T1_0 i	<|special_separator|>
-(056.74, 757.67) (062.49, 757.67) (062.49, 766.42) (056.74, 766.42)      /T1_0 c	<|special_separator|>
-(062.39, 757.67) (065.52, 757.67) (065.52, 766.42) (062.39, 766.42)      /T1_0 l	<|special_separator|>
-(065.42, 757.67) (071.22, 757.67) (071.22, 766.42) (065.42, 766.42)      /T1_0 e	<|special_separator|>
-(391.20, 757.78) (396.23, 757.78) (396.23, 765.05) (391.20, 765.05)      /T1_1 h	<|special_separator|>
-(396.23, 757.78) (399.60, 757.78) (399.60, 765.05) (396.23, 765.05)      /T1_1 t	<|special_separator|>
-(399.60, 757.78) (402.97, 757.78) (402.97, 765.05) (399.60, 765.05)      /T1_1 t	<|special_separator|>
-(402.97, 757.78) (408.09, 757.78) (408.09, 765.05) (402.97, 765.05)      /T1_1 p	<|special_separator|>
-(408.09, 757.78) (411.77, 757.78) (411.77, 765.05) (408.09, 765.05)      /T1_1 s	<|special_separator|>
-(411.77, 757.78) (414.20, 757.78) (414.20, 765.05) (411.77, 765.05)      /T1_1 :	<|special_separator|>
-(414.20, 757.78) (417.33, 757.78) (417.33, 765.05) (414.20, 765.05)      /T1_1 /	<|special_separator|>
-(416.57, 757.78) (419.69, 757.78) (419.69, 765.05) (416.57, 765.05)      /T1_1 /	<|special_separator|>
-(419.21, 757.78) (424.29, 757.78) (424.29, 765.05) (419.21, 765.05)      /T1_1 d	<|special_separator|>
-(424.29, 757.78) (429.25, 757.78) (429.25, 765.05) (424.29, 765.05)      /T1_1 o	<|special_separator|>
-(429.25, 757.78) (431.73, 757.78) (431.73, 765.05) (429.25, 765.05)      /T1_1 i	<|special_separator|>
-(431.73, 757.78) (434.15, 757.78) (434.15, 765.05) (431.73, 765.05)      /T1_1 .	<|special_separator|>
-(433.95, 757.78) (438.90, 757.78) (438.90, 765.05) (433.95, 765.05)      /T1_1 o	<|special_separator|>
-(438.90, 757.78) (442.49, 757.78) (442.49, 765.05) (438.90, 765.05)      /T1_1 r	<|special_separator|>
-(442.41, 757.78) (447.19, 757.78) (447.19, 765.05) (442.41, 765.05)      /T1_1 g	<|special_separator|>
-(447.23, 757.78) (450.36, 757.78) (450.36, 765.05) (447.23, 765.05)      /T1_1 /	<|special_separator|>
-(450.32, 757.78) (453.95, 757.78) (453.95, 765.05) (450.32, 765.05)      /T1_1 1	<|special_separator|>
-(453.95, 757.78) (459.37, 757.78) (459.37, 765.05) (453.95, 765.05)      /T1_1 0	<|special_separator|>
-(458.97, 757.78) (461.40, 757.78) (461.40, 765.05) (458.97, 765.05)      /T1_1 .	<|special_separator|>
-(460.80, 757.78) (464.43, 757.78) (464.43, 765.05) (460.80, 765.05)      /T1_1 1	<|special_separator|>
-(464.43, 757.78) (469.85, 757.78) (469.85, 765.05) (464.43, 765.05)      /T1_1 0	<|special_separator|>
-(469.81, 757.78) (474.29, 757.78) (474.29, 765.05) (469.81, 765.05)      /T1_1 3	<|special_separator|>
-(474.29, 757.78) (479.13, 757.78) (479.13, 765.05) (474.29, 765.05)      /T1_1 8	<|special_separator|>
-(478.93, 757.78) (482.06, 757.78) (482.06, 765.05) (478.93, 765.05)      /T1_1 /	<|special_separator|>
-(481.62, 757.78) (485.31, 757.78) (485.31, 765.05) (481.62, 765.05)      /T1_1 s	<|special_separator|>
-(485.31, 757.78) (490.21, 757.78) (490.21, 765.05) (485.31, 765.05)      /T1_1 4	<|special_separator|>
-(490.21, 757.78) (494.73, 757.78) (494.73, 765.05) (490.21, 765.05)      /T1_1 2	<|special_separator|>
-(494.76, 757.78) (499.28, 757.78) (499.28, 765.05) (494.76, 765.05)      /T1_1 2	<|special_separator|>
-(499.28, 757.78) (503.60, 757.78) (503.60, 765.05) (499.28, 765.05)      /T1_1 5	<|special_separator|>
-(503.52, 757.78) (508.44, 757.78) (508.44, 765.05) (503.52, 765.05)      /T1_1 6	<|special_separator|>
-(508.40, 757.78) (510.92, 757.78) (510.92, 765.05) (508.40, 765.05)      /T1_1 -	<|special_separator|>
-(510.92, 757.78) (516.34, 757.78) (516.34, 765.05) (510.92, 765.05)      /T1_1 0	<|special_separator|>
-(516.34, 757.78) (520.85, 757.78) (520.85, 765.05) (516.34, 765.05)      /T1_1 2	<|special_separator|>
-(520.81, 757.78) (525.73, 757.78) (525.73, 765.05) (520.81, 765.05)      /T1_1 6	<|special_separator|>
-(525.69, 757.78) (528.20, 757.78) (528.20, 765.05) (525.69, 765.05)      /T1_1 -	<|special_separator|>
-(528.20, 757.78) (533.63, 757.78) (533.63, 765.05) (528.20, 765.05)      /T1_1 0	<|special_separator|>
-(533.63, 757.78) (537.26, 757.78) (537.26, 765.05) (533.63, 765.05)      /T1_1 1	<|special_separator|>
-(537.26, 757.78) (541.77, 757.78) (541.77, 765.05) (537.26, 765.05)      /T1_1 2	<|special_separator|>
-(541.57, 757.78) (546.48, 757.78) (546.48, 765.05) (541.57, 765.05)      /T1_1 4	<|special_separator|>
-(546.48, 757.78) (550.99, 757.78) (550.99, 765.05) (546.48, 765.05)      /T1_1 2	<|special_separator|>
-(550.91, 757.78) (553.42, 757.78) (553.42, 765.05) (550.91, 765.05)      /T1_1 -	<|special_separator|>
-(553.42, 757.78) (558.27, 757.78) (558.27, 765.05) (553.42, 765.05)      /T1_1 8	<|special_separator|>
-(039.69, 092.85) (043.70, 092.85) (043.70, 099.21) (039.69, 099.21)      /T1_1 E	<|special_separator|>
-(043.68, 092.85) (047.81, 092.85) (047.81, 099.21) (043.68, 099.21)      /T1_1 x	<|special_separator|>
-(047.65, 092.85) (050.60, 092.85) (050.60, 099.21) (047.65, 099.21)      /T1_1 t	<|special_separator|>
-(050.55, 092.85) (054.35, 092.85) (054.35, 099.21) (050.55, 099.21)      /T1_1 e	<|special_separator|>
-(054.34, 092.85) (058.74, 092.85) (058.74, 099.21) (054.34, 099.21)      /T1_1 n	<|special_separator|>
-(058.72, 092.85) (063.17, 092.85) (063.17, 099.21) (058.72, 099.21)      /T1_1 d	<|special_separator|>
-(063.15, 092.85) (066.95, 092.85) (066.95, 099.21) (063.15, 099.21)      /T1_1 e	<|special_separator|>
-(066.94, 092.85) (071.39, 092.85) (071.39, 099.21) (066.94, 099.21)      /T1_1 d	<|special_separator|>
-(071.37, 092.85) (072.54, 092.85) (072.54, 099.21) (071.37, 099.21)      /T1_1  	<|special_separator|>
-(072.35, 092.85) (077.63, 092.85) (077.63, 099.21) (072.35, 099.21)      /T1_1 D	<|special_separator|>
-(077.62, 092.85) (081.49, 092.85) (081.49, 099.21) (077.62, 099.21)      /T1_1 a	<|special_separator|>
-(081.47, 092.85) (084.42, 092.85) (084.42, 099.21) (081.47, 099.21)      /T1_1 t	<|special_separator|>
-(084.44, 092.85) (088.31, 092.85) (088.31, 099.21) (084.44, 099.21)      /T1_1 a	<|special_separator|>
-(088.30, 092.85) (089.47, 092.85) (089.47, 099.21) (088.30, 099.21)      /T1_1  	<|special_separator|>
-(089.28, 092.85) (093.33, 092.85) (093.33, 099.21) (089.28, 099.21)      /T1_1 F	<|special_separator|>
-(093.24, 092.85) (095.41, 092.85) (095.41, 099.21) (093.24, 099.21)      /T1_1 i	<|special_separator|>
-(095.40, 092.85) (099.59, 092.85) (099.59, 099.21) (095.40, 099.21)      /T1_1 g	<|special_separator|>
-(099.57, 092.85) (101.69, 092.85) (101.69, 099.21) (099.57, 099.21)      /T1_1 .	<|special_separator|>
-(101.68, 092.85) (102.85, 092.85) (102.85, 099.21) (101.68, 099.21)      /T1_1  	<|special_separator|>
-(102.66, 092.85) (106.57, 092.85) (106.57, 099.21) (102.66, 099.21)      /T1_1 3	<|special_separator|>
-(106.56, 092.85) (107.73, 092.85) (107.73, 099.21) (106.56, 099.21)      /T1_1  	<|special_separator|>
-(107.54, 092.85) (109.21, 092.85) (109.21, 099.21) (107.54, 099.21)      /T1_1 |	<|special_separator|>
-(109.20, 092.85) (110.37, 092.85) (110.37, 099.21) (109.20, 099.21)      /T1_1  	<|special_separator|>
-(110.18, 092.85) (114.94, 092.85) (114.94, 099.21) (110.18, 099.21)      /T1_1 V	<|special_separator|>
-(114.47, 092.85) (118.27, 092.85) (118.27, 099.21) (114.47, 099.21)      /T1_1 e	<|special_separator|>
-(118.26, 092.85) (122.65, 092.85) (122.65, 099.21) (118.26, 099.21)      /T1_1 n	<|special_separator|>
-(122.64, 092.85) (127.04, 092.85) (127.04, 099.21) (122.64, 099.21)      /T1_1 n	<|special_separator|>
-(127.02, 092.85) (128.19, 092.85) (128.19, 099.21) (127.02, 099.21)      /T1_1  	<|special_separator|>
-(128.00, 092.85) (132.45, 092.85) (132.45, 099.21) (128.00, 099.21)      /T1_1 d	<|special_separator|>
-(132.43, 092.85) (134.60, 092.85) (134.60, 099.21) (132.43, 099.21)      /T1_1 i	<|special_separator|>
-(134.59, 092.85) (138.46, 092.85) (138.46, 099.21) (134.59, 099.21)      /T1_1 a	<|special_separator|>
-(138.45, 092.85) (142.63, 092.85) (142.63, 099.21) (138.45, 099.21)      /T1_1 g	<|special_separator|>
-(142.62, 092.85) (145.75, 092.85) (145.75, 099.21) (142.62, 099.21)      /T1_1 r	<|special_separator|>
-(145.71, 092.85) (149.58, 092.85) (149.58, 099.21) (145.71, 099.21)      /T1_1 a	<|special_separator|>
-(149.56, 092.85) (156.10, 092.85) (156.10, 099.21) (149.56, 099.21)      /T1_1 m	<|special_separator|>
-(156.08, 092.85) (159.31, 092.85) (159.31, 099.21) (156.08, 099.21)      /T1_1 s	<|special_separator|>
-(159.29, 092.85) (160.46, 092.85) (160.46, 099.21) (159.29, 099.21)      /T1_1  	<|special_separator|>
-(160.27, 092.85) (163.50, 092.85) (163.50, 099.21) (160.27, 099.21)      /T1_1 s	<|special_separator|>
-(163.49, 092.85) (167.89, 092.85) (167.89, 099.21) (163.49, 099.21)      /T1_1 h	<|special_separator|>
-(167.88, 092.85) (172.21, 092.85) (172.21, 099.21) (167.88, 099.21)      /T1_1 o	<|special_separator|>
-(172.02, 092.85) (177.52, 092.85) (177.52, 099.21) (172.02, 099.21)      /T1_1 w	<|special_separator|>
-(177.47, 092.85) (179.64, 092.85) (179.64, 099.21) (177.47, 099.21)      /T1_1 i	<|special_separator|>
-(179.63, 092.85) (184.02, 092.85) (184.02, 099.21) (179.63, 099.21)      /T1_1 n	<|special_separator|>
-(184.01, 092.85) (188.19, 092.85) (188.19, 099.21) (184.01, 099.21)      /T1_1 g	<|special_separator|>
-(188.18, 092.85) (189.35, 092.85) (189.35, 099.21) (188.18, 099.21)      /T1_1  	<|special_separator|>
-(189.16, 092.85) (193.49, 092.85) (193.49, 099.21) (189.16, 099.21)      /T1_1 o	<|special_separator|>
-(193.27, 092.85) (197.16, 092.85) (197.16, 099.21) (193.27, 099.21)      /T1_1 v	<|special_separator|>
-(196.97, 092.85) (200.77, 092.85) (200.77, 099.21) (196.97, 099.21)      /T1_1 e	<|special_separator|>
-(200.76, 092.85) (203.89, 092.85) (203.89, 099.21) (200.76, 099.21)      /T1_1 r	<|special_separator|>
-(203.88, 092.85) (206.08, 092.85) (206.08, 099.21) (203.88, 099.21)      /T1_1 l	<|special_separator|>
-(206.07, 092.85) (209.94, 092.85) (209.94, 099.21) (206.07, 099.21)      /T1_1 a	<|special_separator|>
-(209.93, 092.85) (214.41, 092.85) (214.41, 099.21) (209.93, 099.21)      /T1_1 p	<|special_separator|>
-(214.39, 092.85) (215.56, 092.85) (215.56, 099.21) (214.39, 099.21)      /T1_1  	<|special_separator|>
-(215.37, 092.85) (219.80, 092.85) (219.80, 099.21) (215.37, 099.21)      /T1_1 b	<|special_separator|>
-(219.78, 092.85) (223.58, 092.85) (223.58, 099.21) (219.78, 099.21)      /T1_1 e	<|special_separator|>
-(223.57, 092.85) (226.52, 092.85) (226.52, 099.21) (223.57, 099.21)      /T1_1 t	<|special_separator|>
-(226.50, 092.85) (232.00, 092.85) (232.00, 099.21) (226.50, 099.21)      /T1_1 w	<|special_separator|>
-(231.85, 092.85) (235.65, 092.85) (235.65, 099.21) (231.85, 099.21)      /T1_1 e	<|special_separator|>
-(235.63, 092.85) (239.43, 092.85) (239.43, 099.21) (235.63, 099.21)      /T1_1 e	<|special_separator|>
-(239.42, 092.85) (243.82, 092.85) (243.82, 099.21) (239.42, 099.21)      /T1_1 n	<|special_separator|>
-(243.80, 092.85) (244.97, 092.85) (244.97, 099.21) (243.80, 099.21)      /T1_1  	<|special_separator|>
-(244.78, 092.85) (248.58, 092.85) (248.58, 099.21) (244.78, 099.21)      /T1_1 e	<|special_separator|>
-(248.57, 092.85) (252.97, 092.85) (252.97, 099.21) (248.57, 099.21)      /T1_1 n	<|special_separator|>
-(252.95, 092.85) (256.09, 092.85) (256.09, 099.21) (252.95, 099.21)      /T1_1 r	<|special_separator|>
-(256.07, 092.85) (258.24, 092.85) (258.24, 099.21) (256.07, 099.21)      /T1_1 i	<|special_separator|>
-(258.23, 092.85) (261.86, 092.85) (261.86, 099.21) (258.23, 099.21)      /T1_1 c	<|special_separator|>
-(261.84, 092.85) (266.24, 092.85) (266.24, 099.21) (261.84, 099.21)      /T1_1 h	<|special_separator|>
-(266.23, 092.85) (270.03, 092.85) (270.03, 099.21) (266.23, 099.21)      /T1_1 e	<|special_separator|>
-(270.02, 092.85) (274.46, 092.85) (274.46, 099.21) (270.02, 099.21)      /T1_1 d	<|special_separator|>
-(274.45, 092.85) (275.62, 092.85) (275.62, 099.21) (274.45, 099.21)      /T1_1  	<|special_separator|>
-(275.43, 092.85) (279.91, 092.85) (279.91, 099.21) (275.43, 099.21)      /T1_1 p	<|special_separator|>
-(279.86, 092.85) (283.73, 092.85) (283.73, 099.21) (279.86, 099.21)      /T1_1 a	<|special_separator|>
-(283.72, 092.85) (286.67, 092.85) (286.67, 099.21) (283.72, 099.21)      /T1_1 t	<|special_separator|>
-(286.65, 092.85) (291.06, 092.85) (291.06, 099.21) (286.65, 099.21)      /T1_1 h	<|special_separator|>
-(290.97, 092.85) (296.47, 092.85) (296.47, 099.21) (290.97, 099.21)      /T1_1 w	<|special_separator|>
-(296.35, 092.85) (300.22, 092.85) (300.22, 099.21) (296.35, 099.21)      /T1_1 a	<|special_separator|>
-(300.14, 092.85) (304.09, 092.85) (304.09, 099.21) (300.14, 099.21)      /T1_1 y	<|special_separator|>
-(303.97, 092.85) (307.20, 092.85) (307.20, 099.21) (303.97, 099.21)      /T1_1 s	<|special_separator|>
-(307.19, 092.85) (308.35, 092.85) (308.35, 099.21) (307.19, 099.21)      /T1_1  	<|special_separator|>
-(308.17, 092.85) (310.34, 092.85) (310.34, 099.21) (308.17, 099.21)      /T1_1 i	<|special_separator|>
-(310.32, 092.85) (314.72, 092.85) (314.72, 099.21) (310.32, 099.21)      /T1_1 n	<|special_separator|>
-(314.70, 092.85) (315.87, 092.85) (315.87, 099.21) (314.70, 099.21)      /T1_1  	<|special_separator|>
-(315.68, 092.85) (319.87, 092.85) (319.87, 099.21) (315.68, 099.21)      /T1_1 g	<|special_separator|>
-(319.86, 092.85) (322.99, 092.85) (322.99, 099.21) (319.86, 099.21)      /T1_1 r	<|special_separator|>
-(322.94, 092.85) (327.28, 092.85) (327.28, 099.21) (322.94, 099.21)      /T1_1 o	<|special_separator|>
-(327.26, 092.85) (331.67, 092.85) (331.67, 099.21) (327.26, 099.21)      /T1_1 u	<|special_separator|>
-(331.65, 092.85) (336.05, 092.85) (336.05, 099.21) (331.65, 099.21)      /T1_1 n	<|special_separator|>
-(336.03, 092.85) (340.48, 092.85) (340.48, 099.21) (336.03, 099.21)      /T1_1 d	<|special_separator|>
-(340.47, 092.85) (341.63, 092.85) (341.63, 099.21) (340.47, 099.21)      /T1_1  	<|special_separator|>
-(341.45, 092.85) (344.39, 092.85) (344.39, 099.21) (341.45, 099.21)      /T1_1 t	<|special_separator|>
-(344.38, 092.85) (347.51, 092.85) (347.51, 099.21) (344.38, 099.21)      /T1_1 r	<|special_separator|>
-(347.50, 092.85) (351.90, 092.85) (351.90, 099.21) (347.50, 099.21)      /T1_1 u	<|special_separator|>
-(351.89, 092.85) (354.84, 092.85) (354.84, 099.21) (351.89, 099.21)      /T1_1 t	<|special_separator|>
-(354.82, 092.85) (359.23, 092.85) (359.23, 099.21) (354.82, 099.21)      /T1_1 h	<|special_separator|>
-(359.21, 092.85) (360.38, 092.85) (360.38, 099.21) (359.21, 099.21)      /T1_1  	<|special_separator|>
-(360.19, 092.85) (364.06, 092.85) (364.06, 099.21) (360.19, 099.21)      /T1_1 a	<|special_separator|>
-(364.05, 092.85) (368.44, 092.85) (368.44, 099.21) (364.05, 099.21)      /T1_1 n	<|special_separator|>
-(368.43, 092.85) (372.88, 092.85) (372.88, 099.21) (368.43, 099.21)      /T1_1 d	<|special_separator|>
-(372.86, 092.85) (374.03, 092.85) (374.03, 099.21) (372.86, 099.21)      /T1_1  	<|special_separator|>
-(373.84, 092.85) (378.32, 092.85) (378.32, 099.21) (373.84, 099.21)      /T1_1 p	<|special_separator|>
-(378.31, 092.85) (381.44, 092.85) (381.44, 099.21) (378.31, 099.21)      /T1_1 r	<|special_separator|>
-(381.40, 092.85) (385.20, 092.85) (385.20, 099.21) (381.40, 099.21)      /T1_1 e	<|special_separator|>
-(385.18, 092.85) (389.63, 092.85) (389.63, 099.21) (385.18, 099.21)      /T1_1 d	<|special_separator|>
-(389.62, 092.85) (391.79, 092.85) (391.79, 099.21) (389.62, 099.21)      /T1_1 i	<|special_separator|>
-(391.77, 092.85) (395.40, 092.85) (395.40, 099.21) (391.77, 099.21)      /T1_1 c	<|special_separator|>
-(395.46, 092.85) (398.40, 092.85) (398.40, 099.21) (395.46, 099.21)      /T1_1 t	<|special_separator|>
-(398.36, 092.85) (402.16, 092.85) (402.16, 099.21) (398.36, 099.21)      /T1_1 e	<|special_separator|>
-(402.14, 092.85) (406.59, 092.85) (406.59, 099.21) (402.14, 099.21)      /T1_1 d	<|special_separator|>
-(406.57, 092.85) (407.74, 092.85) (407.74, 099.21) (406.57, 099.21)      /T1_1  	<|special_separator|>
-(407.55, 092.85) (410.78, 092.85) (410.78, 099.21) (407.55, 099.21)      /T1_1 s	<|special_separator|>
-(410.77, 092.85) (412.94, 092.85) (412.94, 099.21) (410.77, 099.21)      /T1_1 i	<|special_separator|>
-(412.92, 092.85) (417.32, 092.85) (417.32, 099.21) (412.92, 099.21)      /T1_1 n	<|special_separator|>
-(417.31, 092.85) (421.49, 092.85) (421.49, 099.21) (417.31, 099.21)      /T1_1 g	<|special_separator|>
-(421.48, 092.85) (423.68, 092.85) (423.68, 099.21) (421.48, 099.21)      /T1_1 l	<|special_separator|>
-(423.67, 092.85) (427.47, 092.85) (427.47, 099.21) (423.67, 099.21)      /T1_1 e	<|special_separator|>
-(427.53, 092.85) (429.73, 092.85) (429.73, 099.21) (427.53, 099.21)      /T1_1 -	<|special_separator|>
-(429.78, 092.85) (433.41, 092.85) (433.41, 099.21) (429.78, 099.21)      /T1_1 c	<|special_separator|>
-(433.40, 092.85) (437.20, 092.85) (437.20, 099.21) (433.40, 099.21)      /T1_1 e	<|special_separator|>
-(437.15, 092.85) (439.36, 092.85) (439.36, 099.21) (437.15, 099.21)      /T1_1 l	<|special_separator|>
-(439.34, 092.85) (441.55, 092.85) (441.55, 099.21) (439.34, 099.21)      /T1_1 l	<|special_separator|>
-(441.53, 092.85) (442.70, 092.85) (442.70, 099.21) (441.53, 099.21)      /T1_1  	<|special_separator|>
-(442.51, 092.85) (446.31, 092.85) (446.31, 099.21) (442.51, 099.21)      /T1_1 e	<|special_separator|>
-(446.16, 092.85) (450.29, 092.85) (450.29, 099.21) (446.16, 099.21)      /T1_1 x	<|special_separator|>
-(450.17, 092.85) (454.65, 092.85) (454.65, 099.21) (450.17, 099.21)      /T1_1 p	<|special_separator|>
-(454.63, 092.85) (457.77, 092.85) (457.77, 099.21) (454.63, 099.21)      /T1_1 r	<|special_separator|>
-(457.72, 092.85) (461.52, 092.85) (461.52, 099.21) (457.72, 099.21)      /T1_1 e	<|special_separator|>
-(461.51, 092.85) (464.74, 092.85) (464.74, 099.21) (461.51, 099.21)      /T1_1 s	<|special_separator|>
-(464.72, 092.85) (467.95, 092.85) (467.95, 099.21) (464.72, 099.21)      /T1_1 s	<|special_separator|>
-(467.94, 092.85) (470.11, 092.85) (470.11, 099.21) (467.94, 099.21)      /T1_1 i	<|special_separator|>
-(470.09, 092.85) (474.43, 092.85) (474.43, 099.21) (470.09, 099.21)      /T1_1 o	<|special_separator|>
-(474.41, 092.85) (478.81, 092.85) (478.81, 099.21) (474.41, 099.21)      /T1_1 n	<|special_separator|>
-(478.79, 092.85) (479.96, 092.85) (479.96, 099.21) (478.79, 099.21)      /T1_1  	<|special_separator|>
-(479.77, 092.85) (482.34, 092.85) (482.34, 099.21) (479.77, 099.21)      /T1_1 f	<|special_separator|>
-(482.25, 092.85) (486.59, 092.85) (486.59, 099.21) (482.25, 099.21)      /T1_1 o	<|special_separator|>
-(486.57, 092.85) (489.71, 092.85) (489.71, 099.21) (486.57, 099.21)      /T1_1 r	<|special_separator|>
-(489.69, 092.85) (490.86, 092.85) (490.86, 099.21) (489.69, 099.21)      /T1_1  	<|special_separator|>
-(490.67, 092.85) (494.55, 092.85) (494.55, 099.21) (490.67, 099.21)      /T1_1 a	<|special_separator|>
-(494.53, 092.85) (496.74, 092.85) (496.74, 099.21) (494.53, 099.21)      /T1_1 l	<|special_separator|>
-(496.72, 092.85) (498.93, 092.85) (498.93, 099.21) (496.72, 099.21)      /T1_1 l	<|special_separator|>
-(498.91, 092.85) (500.08, 092.85) (500.08, 099.21) (498.91, 099.21)      /T1_1  	<|special_separator|>
-(499.89, 092.85) (504.29, 092.85) (504.29, 099.21) (499.89, 099.21)      /T1_1 n	<|special_separator|>
-(504.28, 092.85) (506.45, 092.85) (506.45, 099.21) (504.28, 099.21)      /T1_1 i	<|special_separator|>
-(506.43, 092.85) (510.83, 092.85) (510.83, 099.21) (506.43, 099.21)      /T1_1 n	<|special_separator|>
-(510.81, 092.85) (514.61, 092.85) (514.61, 099.21) (510.81, 099.21)      /T1_1 e	<|special_separator|>
-(514.60, 092.85) (515.77, 092.85) (515.77, 099.21) (514.60, 099.21)      /T1_1  	<|special_separator|>
-(515.58, 092.85) (520.03, 092.85) (520.03, 099.21) (515.58, 099.21)      /T1_1 d	<|special_separator|>
-(520.01, 092.85) (523.15, 092.85) (523.15, 099.21) (520.01, 099.21)      /T1_1 r	<|special_separator|>
-(523.13, 092.85) (527.54, 092.85) (527.54, 099.21) (523.13, 099.21)      /T1_1 u	<|special_separator|>
-(527.52, 092.85) (531.71, 092.85) (531.71, 099.21) (527.52, 099.21)      /T1_1 g	<|special_separator|>
-(531.69, 092.85) (534.92, 092.85) (534.92, 099.21) (531.69, 099.21)      /T1_1 s	<|special_separator|>
-(534.91, 092.85) (536.08, 092.85) (536.08, 099.21) (534.91, 099.21)      /T1_1  	<|special_separator|>
-(535.89, 092.85) (538.06, 092.85) (538.06, 099.21) (535.89, 099.21)      /T1_1 i	<|special_separator|>
-(538.04, 092.85) (542.44, 092.85) (542.44, 099.21) (538.04, 099.21)      /T1_1 n	<|special_separator|>
-(542.44, 092.85) (543.61, 092.85) (543.61, 099.21) (542.44, 099.21)      /T1_1  	<|special_separator|>
-(039.69, 082.85) (042.63, 082.85) (042.63, 089.21) (039.69, 089.21)      /T1_1 t	<|special_separator|>
-(042.62, 082.85) (047.02, 082.85) (047.02, 089.21) (042.62, 089.21)      /T1_1 h	<|special_separator|>
-(047.01, 082.85) (050.81, 082.85) (050.81, 089.21) (047.01, 089.21)      /T1_1 e	<|special_separator|>
-(050.79, 082.85) (051.96, 082.85) (051.96, 089.21) (050.79, 089.21)      /T1_1  	<|special_separator|>
-(051.77, 082.85) (055.77, 082.85) (055.77, 089.21) (051.77, 089.21)      /T1_1 S	<|special_separator|>
-(055.76, 082.85) (059.38, 082.85) (059.38, 089.21) (055.76, 089.21)      /T1_1 c	<|special_separator|>
-(059.37, 082.85) (061.54, 082.85) (061.54, 089.21) (059.37, 089.21)      /T1_1 i	<|special_separator|>
-(061.53, 082.85) (065.90, 082.85) (065.90, 089.21) (061.53, 089.21)      /T1_1 P	<|special_separator|>
-(065.89, 082.85) (068.09, 082.85) (068.09, 089.21) (065.89, 089.21)      /T1_1 l	<|special_separator|>
-(068.08, 082.85) (071.88, 082.85) (071.88, 089.21) (068.08, 089.21)      /T1_1 e	<|special_separator|>
-(071.73, 082.85) (075.85, 082.85) (075.85, 089.21) (071.73, 089.21)      /T1_1 x	<|special_separator|>
-(075.84, 082.85) (077.00, 082.85) (077.00, 089.21) (075.84, 089.21)      /T1_1  	<|special_separator|>
-(076.82, 082.85) (081.26, 082.85) (081.26, 089.21) (076.82, 089.21)      /T1_1 d	<|special_separator|>
-(081.25, 082.85) (085.12, 082.85) (085.12, 089.21) (081.25, 089.21)      /T1_1 a	<|special_separator|>
-(085.10, 082.85) (088.05, 082.85) (088.05, 089.21) (085.10, 089.21)      /T1_1 t	<|special_separator|>
-(088.07, 082.85) (091.94, 082.85) (091.94, 089.21) (088.07, 089.21)      /T1_1 a	<|special_separator|>
-(091.93, 082.85) (095.16, 082.85) (095.16, 089.21) (091.93, 089.21)      /T1_1 s	<|special_separator|>
-(095.14, 082.85) (098.94, 082.85) (098.94, 089.21) (095.14, 089.21)      /T1_1 e	<|special_separator|>
-(098.93, 082.85) (101.88, 082.85) (101.88, 089.21) (098.93, 089.21)      /T1_1 t	<|special_separator|>
-(101.86, 082.85) (103.03, 082.85) (103.03, 089.21) (101.86, 089.21)      /T1_1  	<|special_separator|>
-(102.84, 082.85) (105.34, 082.85) (105.34, 089.21) (102.84, 089.21)      /T1_1 (	<|special_separator|>
-(105.33, 082.85) (107.80, 082.85) (107.80, 089.21) (105.33, 089.21)      /T1_1 I	<|special_separator|>
-(107.79, 082.85) (111.78, 082.85) (111.78, 089.21) (107.79, 089.21)      /T1_1 S	<|special_separator|>
-(111.77, 082.85) (112.94, 082.85) (112.94, 089.21) (111.77, 089.21)      /T1_1  	<|special_separator|>
-(112.75, 082.85) (115.98, 082.85) (115.98, 089.21) (112.75, 089.21)      /T1_1 s	<|special_separator|>
-(115.96, 082.85) (119.76, 082.85) (119.76, 089.21) (115.96, 089.21)      /T1_1 e	<|special_separator|>
-(119.75, 082.85) (122.70, 082.85) (122.70, 089.21) (119.75, 089.21)      /T1_1 t	<|special_separator|>
-(122.65, 082.85) (125.60, 082.85) (125.60, 089.21) (122.65, 089.21)      /T1_1 t	<|special_separator|>
-(125.58, 082.85) (127.75, 082.85) (127.75, 089.21) (125.58, 089.21)      /T1_1 i	<|special_separator|>
-(127.74, 082.85) (132.13, 082.85) (132.13, 089.21) (127.74, 089.21)      /T1_1 n	<|special_separator|>
-(132.12, 082.85) (136.31, 082.85) (136.31, 089.21) (132.12, 089.21)      /T1_1 g	<|special_separator|>
-(136.29, 082.85) (138.45, 082.85) (138.45, 089.21) (136.29, 089.21)      /T1_1 ,	<|special_separator|>
-(138.43, 082.85) (139.60, 082.85) (139.60, 089.21) (138.43, 089.21)      /T1_1  	<|special_separator|>
-(139.41, 082.85) (145.81, 082.85) (145.81, 089.21) (139.41, 089.21)      /T1_1 M	<|special_separator|>
-(145.80, 082.85) (150.13, 082.85) (150.13, 089.21) (145.80, 089.21)      /T1_1 o	<|special_separator|>
-(150.12, 082.85) (155.07, 082.85) (155.07, 089.21) (150.12, 089.21)      /T1_1 A	<|special_separator|>
-(155.05, 082.85) (156.22, 082.85) (156.22, 089.21) (155.05, 089.21)      /T1_1  	<|special_separator|>
-(156.03, 082.85) (159.83, 082.85) (159.83, 089.21) (156.03, 089.21)      /T1_1 e	<|special_separator|>
-(159.82, 082.85) (166.35, 082.85) (166.35, 089.21) (159.82, 089.21)      /T1_1 m	<|special_separator|>
-(166.34, 082.85) (170.76, 082.85) (170.76, 089.21) (166.34, 089.21)      /T1_1 b	<|special_separator|>
-(170.75, 082.85) (174.55, 082.85) (174.55, 089.21) (170.75, 089.21)      /T1_1 e	<|special_separator|>
-(174.53, 082.85) (178.98, 082.85) (178.98, 089.21) (174.53, 089.21)      /T1_1 d	<|special_separator|>
-(178.96, 082.85) (183.41, 082.85) (183.41, 089.21) (178.96, 089.21)      /T1_1 d	<|special_separator|>
-(183.39, 082.85) (185.56, 082.85) (185.56, 089.21) (183.39, 089.21)      /T1_1 i	<|special_separator|>
-(185.55, 082.85) (189.95, 082.85) (189.95, 089.21) (185.55, 089.21)      /T1_1 n	<|special_separator|>
-(189.93, 082.85) (194.12, 082.85) (194.12, 089.21) (189.93, 089.21)      /T1_1 g	<|special_separator|>
-(194.10, 082.85) (197.33, 082.85) (197.33, 089.21) (194.10, 089.21)      /T1_1 s	<|special_separator|>
-(197.25, 082.85) (199.75, 082.85) (199.75, 089.21) (197.25, 089.21)      /T1_1 )	<|special_separator|>
-(199.74, 082.85) (201.86, 082.85) (201.86, 089.21) (199.74, 089.21)      /T1_1 .	<|special_separator|>
-(201.84, 082.71) (203.07, 082.71) (203.07, 089.21) (201.84, 089.21)      /T1_2  	<|special_separator|>
-(202.88, 082.71) (207.52, 082.71) (207.52, 089.21) (202.88, 089.21)      /T1_2 A	<|special_separator|>
-(207.51, 082.71) (208.74, 082.71) (208.74, 089.21) (207.51, 089.21)      /T1_2  	<|special_separator|>
-(208.54, 082.71) (211.22, 082.71) (211.22, 089.21) (208.54, 089.21)      /T1_2 t	<|special_separator|>
-(211.18, 082.71) (215.41, 082.71) (215.41, 089.21) (211.18, 089.21)      /T1_2 o	<|special_separator|>
-(215.40, 082.71) (218.08, 082.71) (218.08, 089.21) (215.40, 089.21)      /T1_2 t	<|special_separator|>
-(218.06, 082.71) (221.69, 082.71) (221.69, 089.21) (218.06, 089.21)      /T1_2 a	<|special_separator|>
-(221.68, 082.71) (223.67, 082.71) (223.67, 089.21) (221.68, 089.21)      /T1_2 l	<|special_separator|>
-(223.66, 082.71) (224.89, 082.71) (224.89, 089.21) (223.66, 089.21)      /T1_2  	<|special_separator|>
-(224.69, 082.71) (228.93, 082.71) (228.93, 089.21) (224.69, 089.21)      /T1_2 o	<|special_separator|>
-(228.88, 082.71) (231.41, 082.71) (231.41, 089.21) (228.88, 089.21)      /T1_2 f	<|special_separator|>
-(231.39, 082.71) (232.63, 082.71) (232.63, 089.21) (231.39, 089.21)      /T1_2  	<|special_separator|>
-(232.43, 082.71) (236.61, 082.71) (236.61, 089.21) (232.43, 089.21)      /T1_2 n	<|special_separator|>
-(236.59, 082.71) (240.64, 082.71) (240.64, 089.21) (236.59, 089.21)      /T1_2 =	<|special_separator|>
-(240.63, 082.71) (244.81, 082.71) (244.81, 089.21) (240.63, 089.21)      /T1_2 6	<|special_separator|>
-(244.72, 082.71) (248.31, 082.71) (248.31, 089.21) (244.72, 089.21)      /T1_2 5	<|special_separator|>
-(248.29, 082.71) (252.12, 082.71) (252.12, 089.21) (248.29, 089.21)      /T1_2 3	<|special_separator|>
-(252.11, 082.71) (253.34, 082.71) (253.34, 089.21) (252.11, 089.21)      /T1_2  	<|special_separator|>
-(253.14, 082.71) (257.47, 082.71) (257.47, 089.21) (253.14, 089.21)      /T1_2 p	<|special_separator|>
-(257.39, 082.71) (261.01, 082.71) (261.01, 089.21) (257.39, 089.21)      /T1_2 a	<|special_separator|>
-(261.00, 082.71) (263.68, 082.71) (263.68, 089.21) (261.00, 089.21)      /T1_2 t	<|special_separator|>
-(263.67, 082.71) (267.85, 082.71) (267.85, 089.21) (263.67, 089.21)      /T1_2 h	<|special_separator|>
-(267.81, 082.71) (273.07, 082.71) (273.07, 089.21) (267.81, 089.21)      /T1_2 w	<|special_separator|>
-(273.02, 082.71) (276.65, 082.71) (276.65, 089.21) (273.02, 089.21)      /T1_2 a	<|special_separator|>
-(276.60, 082.71) (280.28, 082.71) (280.28, 089.21) (276.60, 089.21)      /T1_2 y	<|special_separator|>
-(280.24, 082.71) (283.41, 082.71) (283.41, 089.21) (280.24, 089.21)      /T1_2 s	<|special_separator|>
-(283.39, 082.71) (284.62, 082.71) (284.62, 089.21) (283.39, 089.21)      /T1_2  	<|special_separator|>
-(284.43, 082.71) (289.69, 082.71) (289.69, 089.21) (284.43, 089.21)      /T1_2 w	<|special_separator|>
-(289.61, 082.71) (293.33, 082.71) (293.33, 089.21) (289.61, 089.21)      /T1_2 e	<|special_separator|>
-(293.32, 082.71) (296.23, 082.71) (296.23, 089.21) (293.32, 089.21)      /T1_2 r	<|special_separator|>
-(296.11, 082.71) (299.83, 082.71) (299.83, 089.21) (296.11, 089.21)      /T1_2 e	<|special_separator|>
-(299.82, 082.71) (301.05, 082.71) (301.05, 089.21) (299.82, 089.21)      /T1_2  	<|special_separator|>
-(300.85, 082.71) (303.54, 082.71) (303.54, 089.21) (300.85, 089.21)      /T1_2 t	<|special_separator|>
-(303.45, 082.71) (307.18, 082.71) (307.18, 089.21) (303.45, 089.21)      /T1_2 e	<|special_separator|>
-(307.16, 082.71) (310.33, 082.71) (310.33, 089.21) (307.16, 089.21)      /T1_2 s	<|special_separator|>
-(310.29, 082.71) (312.97, 082.71) (312.97, 089.21) (310.29, 089.21)      /T1_2 t	<|special_separator|>
-(312.89, 082.71) (316.61, 082.71) (316.61, 089.21) (312.89, 089.21)      /T1_2 e	<|special_separator|>
-(316.60, 082.71) (320.90, 082.71) (320.90, 089.21) (316.60, 089.21)      /T1_2 d	<|special_separator|>
-(320.89, 082.71) (322.78, 082.71) (322.78, 089.21) (320.89, 089.21)      /T1_2 .	<|special_separator|>
-(322.76, 082.71) (323.99, 082.71) (323.99, 089.21) (322.76, 089.21)      /T1_2  	<|special_separator|>
-(323.80, 082.71) (327.63, 082.71) (327.63, 089.21) (323.80, 089.21)      /T1_2 F	<|special_separator|>
-(327.41, 082.71) (331.64, 082.71) (331.64, 089.21) (327.41, 089.21)      /T1_2 o	<|special_separator|>
-(331.63, 082.71) (334.53, 082.71) (334.53, 089.21) (331.63, 089.21)      /T1_2 r	<|special_separator|>
-(334.52, 082.71) (335.75, 082.71) (335.75, 089.21) (334.52, 089.21)      /T1_2  	<|special_separator|>
-(335.55, 082.71) (339.18, 082.71) (339.18, 089.21) (335.55, 089.21)      /T1_2 a	<|special_separator|>
-(339.17, 082.71) (341.16, 082.71) (341.16, 089.21) (339.17, 089.21)      /T1_2 l	<|special_separator|>
-(341.15, 082.71) (343.14, 082.71) (343.14, 089.21) (341.15, 089.21)      /T1_2 l	<|special_separator|>
-(343.13, 082.71) (344.36, 082.71) (344.36, 089.21) (343.13, 089.21)      /T1_2  	<|special_separator|>
-(344.16, 082.71) (348.47, 082.71) (348.47, 089.21) (344.16, 089.21)      /T1_2 d	<|special_separator|>
-(348.46, 082.71) (351.36, 082.71) (351.36, 089.21) (348.46, 089.21)      /T1_2 r	<|special_separator|>
-(351.35, 082.71) (355.54, 082.71) (355.54, 089.21) (351.35, 089.21)      /T1_2 u	<|special_separator|>
-(355.53, 082.71) (359.60, 082.71) (359.60, 089.21) (355.53, 089.21)      /T1_2 g	<|special_separator|>
-(359.55, 082.71) (362.72, 082.71) (362.72, 089.21) (359.55, 089.21)      /T1_2 s	<|special_separator|>
-(362.71, 082.71) (364.63, 082.71) (364.63, 089.21) (362.71, 089.21)      /T1_2 ,	<|special_separator|>
-(364.62, 082.71) (365.85, 082.71) (365.85, 089.21) (364.62, 089.21)      /T1_2  	<|special_separator|>
-(365.66, 082.71) (368.34, 082.71) (368.34, 089.21) (365.66, 089.21)      /T1_2 t	<|special_separator|>
-(368.32, 082.71) (372.51, 082.71) (372.51, 089.21) (368.32, 089.21)      /T1_2 h	<|special_separator|>
-(372.49, 082.71) (376.22, 082.71) (376.22, 089.21) (372.49, 089.21)      /T1_2 e	<|special_separator|>
-(376.20, 082.71) (377.44, 082.71) (377.44, 089.21) (376.20, 089.21)      /T1_2  	<|special_separator|>
-(377.24, 082.71) (381.42, 082.71) (381.42, 089.21) (377.24, 089.21)      /T1_2 n	<|special_separator|>
-(381.41, 082.71) (385.60, 082.71) (385.60, 089.21) (381.41, 089.21)      /T1_2 u	<|special_separator|>
-(385.58, 082.71) (391.92, 082.71) (391.92, 089.21) (385.58, 089.21)      /T1_2 m	<|special_separator|>
-(391.91, 082.71) (396.15, 082.71) (396.15, 089.21) (391.91, 089.21)      /T1_2 b	<|special_separator|>
-(396.14, 082.71) (399.86, 082.71) (399.86, 089.21) (396.14, 089.21)      /T1_2 e	<|special_separator|>
-(399.85, 082.71) (402.76, 082.71) (402.76, 089.21) (399.85, 089.21)      /T1_2 r	<|special_separator|>
-(402.74, 082.71) (403.97, 082.71) (403.97, 089.21) (402.74, 089.21)      /T1_2  	<|special_separator|>
-(403.78, 082.71) (408.01, 082.71) (408.01, 089.21) (403.78, 089.21)      /T1_2 o	<|special_separator|>
-(407.96, 082.71) (410.49, 082.71) (410.49, 089.21) (407.96, 089.21)      /T1_2 f	<|special_separator|>
-(410.48, 082.71) (411.71, 082.71) (411.71, 089.21) (410.48, 089.21)      /T1_2  	<|special_separator|>
-(411.51, 082.71) (415.10, 082.71) (415.10, 089.21) (411.51, 089.21)      /T1_2 c	<|special_separator|>
-(415.05, 082.71) (419.29, 082.71) (419.29, 089.21) (415.05, 089.21)      /T1_2 o	<|special_separator|>
-(419.27, 082.71) (425.61, 082.71) (425.61, 089.21) (419.27, 089.21)      /T1_2 m	<|special_separator|>
-(425.59, 082.71) (431.93, 082.71) (431.93, 089.21) (425.59, 089.21)      /T1_2 m	<|special_separator|>
-(431.91, 082.71) (436.15, 082.71) (436.15, 089.21) (431.91, 089.21)      /T1_2 o	<|special_separator|>
-(436.13, 082.71) (440.31, 082.71) (440.31, 089.21) (436.13, 089.21)      /T1_2 n	<|special_separator|>
-(440.30, 082.71) (441.53, 082.71) (441.53, 089.21) (440.30, 089.21)      /T1_2  	<|special_separator|>
-(441.34, 082.71) (445.06, 082.71) (445.06, 089.21) (441.34, 089.21)      /T1_2 e	<|special_separator|>
-(445.05, 082.71) (449.22, 082.71) (449.22, 089.21) (445.05, 089.21)      /T1_2 n	<|special_separator|>
-(449.21, 082.71) (452.12, 082.71) (452.12, 089.21) (449.21, 089.21)      /T1_2 r	<|special_separator|>
-(452.10, 082.71) (454.10, 082.71) (454.10, 089.21) (452.10, 089.21)      /T1_2 i	<|special_separator|>
-(454.09, 082.71) (457.67, 082.71) (457.67, 089.21) (454.09, 089.21)      /T1_2 c	<|special_separator|>
-(457.66, 082.71) (461.85, 082.71) (461.85, 089.21) (457.66, 089.21)      /T1_2 h	<|special_separator|>
-(461.83, 082.71) (465.56, 082.71) (465.56, 089.21) (461.83, 089.21)      /T1_2 e	<|special_separator|>
-(465.54, 082.71) (469.85, 082.71) (469.85, 089.21) (465.54, 089.21)      /T1_2 d	<|special_separator|>
-(469.83, 082.71) (471.06, 082.71) (471.06, 089.21) (469.83, 089.21)      /T1_2  	<|special_separator|>
-(470.87, 082.71) (475.19, 082.71) (475.19, 089.21) (470.87, 089.21)      /T1_2 p	<|special_separator|>
-(475.11, 082.71) (478.74, 082.71) (478.74, 089.21) (475.11, 089.21)      /T1_2 a	<|special_separator|>
-(478.72, 082.71) (481.41, 082.71) (481.41, 089.21) (478.72, 089.21)      /T1_2 t	<|special_separator|>
-(481.39, 082.71) (485.58, 082.71) (485.58, 089.21) (481.39, 089.21)      /T1_2 h	<|special_separator|>
-(485.53, 082.71) (490.79, 082.71) (490.79, 089.21) (485.53, 089.21)      /T1_2 w	<|special_separator|>
-(490.75, 082.71) (494.37, 082.71) (494.37, 089.21) (490.75, 089.21)      /T1_2 a	<|special_separator|>
-(494.33, 082.71) (498.01, 082.71) (498.01, 089.21) (494.33, 089.21)      /T1_2 y	<|special_separator|>
-(497.96, 082.71) (501.13, 082.71) (501.13, 089.21) (497.96, 089.21)      /T1_2 s	<|special_separator|>
-(501.12, 082.71) (502.35, 082.71) (502.35, 089.21) (501.12, 089.21)      /T1_2  	<|special_separator|>
-(502.15, 082.71) (507.42, 082.71) (507.42, 089.21) (502.15, 089.21)      /T1_2 w	<|special_separator|>
-(507.37, 082.71) (511.00, 082.71) (511.00, 089.21) (507.37, 089.21)      /T1_2 a	<|special_separator|>
-(510.98, 082.71) (514.15, 082.71) (514.15, 089.21) (510.98, 089.21)      /T1_2 s	<|special_separator|>
-(514.14, 082.71) (515.37, 082.71) (515.37, 089.21) (514.14, 089.21)      /T1_2  	<|special_separator|>
-(515.17, 082.71) (518.35, 082.71) (518.35, 089.21) (515.17, 089.21)      /T1_2 s	<|special_separator|>
-(518.30, 082.71) (520.98, 082.71) (520.98, 089.21) (518.30, 089.21)      /T1_2 t	<|special_separator|>
-(520.97, 082.71) (524.59, 082.71) (524.59, 089.21) (520.97, 089.21)      /T1_2 a	<|special_separator|>
-(524.58, 082.71) (527.26, 082.71) (527.26, 089.21) (524.58, 089.21)      /T1_2 t	<|special_separator|>
-(527.24, 082.71) (529.25, 082.71) (529.25, 089.21) (527.24, 089.21)      /T1_2 i	<|special_separator|>
-(529.23, 082.71) (532.40, 082.71) (532.40, 089.21) (529.23, 089.21)      /T1_2 s	<|special_separator|>
-(532.36, 082.71) (535.04, 082.71) (535.04, 089.21) (532.36, 089.21)      /T1_2 t	<|special_separator|>
-(535.02, 082.71) (537.02, 082.71) (537.02, 089.21) (535.02, 089.21)      /T1_2 i	<|special_separator|>
-(537.01, 082.71) (540.59, 082.71) (540.59, 089.21) (537.01, 089.21)      /T1_2 c	<|special_separator|>
-(540.58, 082.71) (544.21, 082.71) (544.21, 089.21) (540.58, 089.21)      /T1_2 a	<|special_separator|>
-(544.19, 082.71) (546.19, 082.71) (546.19, 089.21) (544.19, 089.21)      /T1_2 l	<|special_separator|>
-(546.17, 082.71) (548.17, 082.71) (548.17, 089.21) (546.17, 089.21)      /T1_2 l	<|special_separator|>
-(548.15, 082.71) (551.84, 082.71) (551.84, 089.21) (548.15, 089.21)      /T1_2 y	<|special_separator|>
-(551.84, 082.71) (553.07, 082.71) (553.07, 089.21) (551.84, 089.21)      /T1_2  	<|special_separator|>
-(039.69, 072.71) (042.86, 072.71) (042.86, 079.21) (039.69, 079.21)      /T1_2 s	<|special_separator|>
-(042.84, 072.71) (044.84, 072.71) (044.84, 079.21) (042.84, 079.21)      /T1_2 i	<|special_separator|>
-(044.83, 072.71) (048.90, 072.71) (048.90, 079.21) (044.83, 079.21)      /T1_2 g	<|special_separator|>
-(048.89, 072.71) (053.07, 072.71) (053.07, 079.21) (048.89, 079.21)      /T1_2 n	<|special_separator|>
-(053.06, 072.71) (055.06, 072.71) (055.06, 079.21) (053.06, 079.21)      /T1_2 i	<|special_separator|>
-(055.04, 072.71) (057.57, 072.71) (057.57, 079.21) (055.04, 079.21)      /T1_2 f	<|special_separator|>
-(057.63, 072.71) (059.63, 072.71) (059.63, 079.21) (057.63, 079.21)      /T1_2 i	<|special_separator|>
-(059.62, 072.71) (063.20, 072.71) (063.20, 079.21) (059.62, 079.21)      /T1_2 c	<|special_separator|>
-(063.19, 072.71) (066.81, 072.71) (066.81, 079.21) (063.19, 079.21)      /T1_2 a	<|special_separator|>
-(066.80, 072.71) (070.98, 072.71) (070.98, 079.21) (066.80, 079.21)      /T1_2 n	<|special_separator|>
-(070.96, 072.71) (073.64, 072.71) (073.64, 079.21) (070.96, 079.21)      /T1_2 t	<|special_separator|>
-(073.63, 072.71) (075.52, 072.71) (075.52, 079.21) (073.63, 079.21)      /T1_2 .	<|special_separator|>
-(075.51, 072.71) (076.74, 072.71) (076.74, 079.21) (075.51, 079.21)      /T1_2  	<|special_separator|>
-(076.54, 072.71) (081.90, 072.71) (081.90, 079.21) (076.54, 079.21)      /T1_2 O	<|special_separator|>
-(081.89, 072.71) (086.07, 072.71) (086.07, 079.21) (081.89, 079.21)      /T1_2 n	<|special_separator|>
-(086.06, 072.71) (089.78, 072.71) (089.78, 079.21) (086.06, 079.21)      /T1_2 e	<|special_separator|>
-(089.77, 072.71) (091.92, 072.71) (091.92, 079.21) (089.77, 079.21)      /T1_2 -	<|special_separator|>
-(091.91, 072.71) (094.59, 072.71) (094.59, 079.21) (091.91, 079.21)      /T1_2 t	<|special_separator|>
-(094.57, 072.71) (098.20, 072.71) (098.20, 079.21) (094.57, 079.21)      /T1_2 a	<|special_separator|>
-(098.19, 072.71) (100.19, 072.71) (100.19, 079.21) (098.19, 079.21)      /T1_2 i	<|special_separator|>
-(100.17, 072.71) (102.17, 072.71) (102.17, 079.21) (100.17, 079.21)      /T1_2 l	<|special_separator|>
-(102.16, 072.71) (105.88, 072.71) (105.88, 079.21) (102.16, 079.21)      /T1_2 e	<|special_separator|>
-(105.87, 072.71) (110.17, 072.71) (110.17, 079.21) (105.87, 079.21)      /T1_2 d	<|special_separator|>
-(110.16, 072.71) (111.39, 072.71) (111.39, 079.21) (110.16, 079.21)      /T1_2  	<|special_separator|>
-(111.19, 072.71) (115.38, 072.71) (115.38, 079.21) (111.19, 079.21)      /T1_2 h	<|special_separator|>
-(115.33, 072.71) (119.01, 072.71) (119.01, 079.21) (115.33, 079.21)      /T1_2 y	<|special_separator|>
-(119.00, 072.71) (123.32, 072.71) (123.32, 079.21) (119.00, 079.21)      /T1_2 p	<|special_separator|>
-(123.31, 072.71) (127.03, 072.71) (127.03, 079.21) (123.31, 079.21)      /T1_2 e	<|special_separator|>
-(127.02, 072.71) (129.93, 072.71) (129.93, 079.21) (127.02, 079.21)      /T1_2 r	<|special_separator|>
-(129.84, 072.71) (133.92, 072.71) (133.92, 079.21) (129.84, 079.21)      /T1_2 g	<|special_separator|>
-(133.76, 072.71) (137.49, 072.71) (137.49, 079.21) (133.76, 079.21)      /T1_2 e	<|special_separator|>
-(137.47, 072.71) (141.71, 072.71) (141.71, 079.21) (137.47, 079.21)      /T1_2 o	<|special_separator|>
-(141.70, 072.71) (148.03, 072.71) (148.03, 079.21) (141.70, 079.21)      /T1_2 m	<|special_separator|>
-(148.02, 072.71) (151.74, 072.71) (151.74, 079.21) (148.02, 079.21)      /T1_2 e	<|special_separator|>
-(151.73, 072.71) (154.41, 072.71) (154.41, 079.21) (151.73, 079.21)      /T1_2 t	<|special_separator|>
-(154.39, 072.71) (157.30, 072.71) (157.30, 079.21) (154.39, 079.21)      /T1_2 r	<|special_separator|>
-(157.28, 072.71) (159.29, 072.71) (159.29, 079.21) (157.28, 079.21)      /T1_2 i	<|special_separator|>
-(159.27, 072.71) (162.86, 072.71) (162.86, 079.21) (159.27, 079.21)      /T1_2 c	<|special_separator|>
-(162.84, 072.71) (164.07, 072.71) (164.07, 079.21) (162.84, 079.21)      /T1_2  	<|special_separator|>
-(163.88, 072.71) (168.20, 072.71) (168.20, 079.21) (163.88, 079.21)      /T1_2 p	<|special_separator|>
-(168.23, 072.71) (170.38, 072.71) (170.38, 079.21) (168.23, 079.21)      /T1_2 -	<|special_separator|>
-(170.34, 072.71) (174.05, 072.71) (174.05, 079.21) (170.34, 079.21)      /T1_2 v	<|special_separator|>
-(173.97, 072.71) (177.60, 072.71) (177.60, 079.21) (173.97, 079.21)      /T1_2 a	<|special_separator|>
-(177.58, 072.71) (179.58, 072.71) (179.58, 079.21) (177.58, 079.21)      /T1_2 l	<|special_separator|>
-(179.56, 072.71) (183.76, 072.71) (183.76, 079.21) (179.56, 079.21)      /T1_2 u	<|special_separator|>
-(183.74, 072.71) (187.47, 072.71) (187.47, 079.21) (183.74, 079.21)      /T1_2 e	<|special_separator|>
-(187.45, 072.71) (188.68, 072.71) (188.68, 079.21) (187.45, 079.21)      /T1_2  	<|special_separator|>
-(188.49, 072.71) (192.81, 072.71) (192.81, 079.21) (188.49, 079.21)      /T1_2 p	<|special_separator|>
-(192.80, 072.71) (196.83, 072.71) (196.83, 079.21) (192.80, 079.21)      /T1_2 <	<|special_separator|>
-(196.82, 072.71) (199.74, 072.71) (199.74, 079.21) (196.82, 079.21)      /T1_2 1	<|special_separator|>
-(199.73, 072.71) (203.45, 072.71) (203.45, 079.21) (199.73, 079.21)      /T1_2 e	<|special_separator|>
-(203.44, 072.71) (205.60, 072.71) (205.60, 079.21) (203.44, 079.21)      /T1_2 -	<|special_separator|>
-(205.58, 072.71) (208.51, 072.71) (208.51, 079.21) (205.58, 079.21)      /T1_2 1	<|special_separator|>
-(208.49, 072.71) (212.32, 072.71) (212.32, 079.21) (208.49, 079.21)      /T1_2 3	<|special_separator|>
-(212.31, 072.71) (213.54, 072.71) (213.54, 079.21) (212.31, 079.21)      /T1_2  	<|special_separator|>
-(213.35, 072.71) (215.35, 072.71) (215.35, 079.21) (213.35, 079.21)      /T1_2 i	<|special_separator|>
-(215.33, 072.71) (219.51, 072.71) (219.51, 079.21) (215.33, 079.21)      /T1_2 n	<|special_separator|>
-(219.50, 072.71) (220.73, 072.71) (220.73, 079.21) (219.50, 079.21)      /T1_2  	<|special_separator|>
-(220.53, 072.71) (224.16, 072.71) (224.16, 079.21) (220.53, 079.21)      /T1_2 a	<|special_separator|>
-(224.15, 072.71) (226.14, 072.71) (226.14, 079.21) (224.15, 079.21)      /T1_2 l	<|special_separator|>
-(226.13, 072.71) (228.12, 072.71) (228.12, 079.21) (226.13, 079.21)      /T1_2 l	<|special_separator|>
-(228.11, 072.71) (229.34, 072.71) (229.34, 079.21) (228.11, 079.21)      /T1_2  	<|special_separator|>
-(229.14, 072.71) (232.73, 072.71) (232.73, 079.21) (229.14, 079.21)      /T1_2 c	<|special_separator|>
-(232.71, 072.71) (236.34, 072.71) (236.34, 079.21) (232.71, 079.21)      /T1_2 a	<|special_separator|>
-(236.33, 072.71) (239.50, 072.71) (239.50, 079.21) (236.33, 079.21)      /T1_2 s	<|special_separator|>
-(239.41, 072.71) (243.14, 072.71) (243.14, 079.21) (239.41, 079.21)      /T1_2 e	<|special_separator|>
-(243.12, 072.71) (246.30, 072.71) (246.30, 079.21) (243.12, 079.21)      /T1_2 s	<|special_separator|>
-(246.28, 072.71) (248.17, 072.71) (248.17, 079.21) (246.28, 079.21)      /T1_2 .

groundtruth/parser/font_11.pdf.page_no_16.py.json.line.txt

@@ -1,7 +0,0 @@
-(039.69, 020.40) (145.28, 020.40) (145.28, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 092.85) (543.61, 092.85) (543.61, 099.21) (039.69, 099.21)      /T1_1 Extended Data Fig. 3 | Venn diagrams showing overlap between enriched pathways in ground truth and predicted single-cell expression for all nine drugs in 	<|special_separator|>
-(039.69, 082.85) (201.86, 082.85) (201.86, 089.21) (039.69, 089.21)      /T1_1 the SciPlex dataset (IS setting, MoA embeddings).	<|special_separator|>
-(201.84, 082.71) (553.07, 082.71) (553.07, 089.21) (201.84, 089.21)      /T1_2  A total of n=653 pathways were tested. For all drugs, the number of common enriched pathways was statistically 	<|special_separator|>
-(039.69, 072.71) (248.17, 072.71) (248.17, 079.21) (039.69, 079.21)      /T1_2 significant. One-tailed hypergeometric p-value p<1e-13 in all cases.

groundtruth/parser/font_11.pdf.page_no_16.py.json.word.txt

@@ -1,8 +0,0 @@
-(039.69, 020.40) (065.17, 020.40) (065.17, 027.41) (039.69, 027.41)      /T1_0 Nature	<|special_separator|>
-(066.63, 020.40) (145.28, 020.40) (145.28, 027.41) (066.63, 027.41)      /T1_0 MachineIntelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 092.85) (542.44, 092.85) (542.44, 099.21) (039.69, 099.21)      /T1_1 ExtendedDataFig.3|Venndiagramsshowingoverlapbetweenenrichedpathwaysingroundtruthandpredictedsingle-cellexpressionforallninedrugsin	<|special_separator|>
-(039.69, 082.85) (201.86, 082.85) (201.86, 089.21) (039.69, 089.21)      /T1_1 theSciPlexdataset(ISsetting,MoAembeddings).	<|special_separator|>
-(202.88, 082.71) (551.84, 082.71) (551.84, 089.21) (202.88, 089.21)      /T1_2 Atotalofn=653pathwaysweretested.Foralldrugs,thenumberofcommonenrichedpathwayswasstatistically	<|special_separator|>
-(039.69, 072.71) (248.17, 072.71) (248.17, 079.21) (039.69, 079.21)      /T1_2 significant.One-tailedhypergeometricp-valuep<1e-13inallcases.

groundtruth/parser/font_11.pdf.page_no_17.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a7c65993ea15563ef1cd99174df3f5f049c48823631ca01898daf463a369107a
-size 1460450

groundtruth/parser/font_11.pdf.page_no_17.py.json.line.txt

@@ -1,21 +0,0 @@
-(039.69, 020.40) (145.28, 020.40) (145.28, 027.41) (039.69, 027.41)      /T1_0 Nature Machine Intelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 482.85) (257.96, 482.85) (257.96, 489.21) (039.69, 489.21)      /T1_1 Extended Data Fig. 4 | Comparison of different Monge, Conditional 	<|special_separator|>
-(039.69, 472.85) (275.25, 472.85) (275.25, 479.21) (039.69, 479.21)      /T1_1 Monge (CMonge), within condition, and identity models using the mean 	<|special_separator|>
-(039.69, 462.85) (112.08, 462.85) (112.08, 469.21) (039.69, 469.21)      /T1_1 Wasserstein Distance.	<|special_separator|>
-(112.06, 462.71) (293.95, 462.71) (293.95, 469.21) (112.06, 469.21)      /T1_2  Panels show overall performance, performance for single 	<|special_separator|>
-(039.69, 452.71) (282.40, 452.71) (282.40, 459.20) (039.69, 459.20)      /T1_2 treatments, and combinatorial treatments. Boxes indicate the three quartiles 	<|special_separator|>
-(039.69, 442.70) (279.14, 442.70) (279.14, 449.20) (039.69, 449.20)      /T1_2 (25th percentile, median and 75th percentile). The whiskers extend to points 	<|special_separator|>
-(039.69, 432.70) (284.86, 432.70) (284.86, 439.20) (039.69, 439.20)      /T1_2 that lie within 1.5 times the interquartile range of the lower and upper quartile. 	<|special_separator|>
-(039.69, 422.84) (044.63, 422.84) (044.63, 429.19) (039.69, 429.19)      /T1_1 A	<|special_separator|>
-(044.62, 422.70) (274.12, 422.70) (274.12, 429.19) (044.62, 429.19)      /T1_2 ) In-sample results. The Monge model is trained as one model that sees all 	<|special_separator|>
-(306.47, 482.69) (556.96, 482.69) (556.96, 489.19) (306.47, 489.19)      /T1_2 conditions but is not aware of conditional information. The CMonge models are 	<|special_separator|>
-(306.47, 472.69) (548.17, 472.69) (548.17, 479.19) (306.47, 479.19)      /T1_2 likewise trained on all conditions but do get conditional information with the 	<|special_separator|>
-(306.47, 462.69) (446.29, 462.69) (446.29, 469.19) (306.47, 469.19)      /T1_2 Mode-of-Action (MoA) or RDkit embedding. 	<|special_separator|>
-(446.09, 462.83) (450.79, 462.83) (450.79, 469.19) (446.09, 469.19)      /T1_1 B	<|special_separator|>
-(450.82, 462.69) (547.15, 462.69) (547.15, 469.19) (450.82, 469.19)      /T1_2 ) Out-of-sample (OOS) results. 	<|special_separator|>
-(306.47, 452.69) (543.15, 452.69) (543.15, 459.18) (306.47, 459.18)      /T1_2 All models are trained in a leave-one-drug-out setting. This means that each 	<|special_separator|>
-(306.47, 442.68) (554.35, 442.68) (554.35, 449.18) (306.47, 449.18)      /T1_2 boxenplot shows the result over 33 drugs, from 33 models. The Monge model is 	<|special_separator|>
-(306.47, 432.68) (560.43, 432.68) (560.43, 439.18) (306.47, 439.18)      /T1_2 also trained in a leave-one-drug-out setting but does not incorporate conditional 	<|special_separator|>
-(306.47, 422.68) (347.11, 422.68) (347.11, 429.17) (306.47, 429.17)      /T1_2 information.

groundtruth/parser/font_11.pdf.page_no_17.py.json.word.txt

@@ -1,22 +0,0 @@
-(039.69, 020.40) (065.17, 020.40) (065.17, 027.41) (039.69, 027.41)      /T1_0 Nature	<|special_separator|>
-(066.63, 020.40) (145.28, 020.40) (145.28, 027.41) (066.63, 027.41)      /T1_0 MachineIntelligence	<|special_separator|>
-(039.69, 757.67) (071.22, 757.67) (071.22, 766.42) (039.69, 766.42)      /T1_0 Article	<|special_separator|>
-(391.20, 757.78) (558.27, 757.78) (558.27, 765.05) (391.20, 765.05)      /T1_1 https://doi.org/10.1038/s42256-026-01242-8	<|special_separator|>
-(039.69, 482.85) (256.79, 482.85) (256.79, 489.21) (039.69, 489.21)      /T1_1 ExtendedDataFig.4|ComparisonofdifferentMonge,Conditional	<|special_separator|>
-(039.69, 472.85) (274.08, 472.85) (274.08, 479.21) (039.69, 479.21)      /T1_1 Monge(CMonge),withincondition,andidentitymodelsusingthemean	<|special_separator|>
-(039.69, 462.85) (112.08, 462.85) (112.08, 469.21) (039.69, 469.21)      /T1_1 WassersteinDistance.	<|special_separator|>
-(113.10, 462.71) (292.71, 462.71) (292.71, 469.21) (113.10, 469.21)      /T1_2 Panelsshowoverallperformance,performanceforsingle	<|special_separator|>
-(039.69, 452.71) (281.17, 452.71) (281.17, 459.20) (039.69, 459.20)      /T1_2 treatments,andcombinatorialtreatments.Boxesindicatethethreequartiles	<|special_separator|>
-(039.69, 442.70) (277.91, 442.70) (277.91, 449.20) (039.69, 449.20)      /T1_2 (25thpercentile,medianand75thpercentile).Thewhiskersextendtopoints	<|special_separator|>
-(039.69, 432.70) (283.63, 432.70) (283.63, 439.20) (039.69, 439.20)      /T1_2 thatliewithin1.5timestheinterquartilerangeofthelowerandupperquartile.	<|special_separator|>
-(039.69, 422.84) (044.63, 422.84) (044.63, 429.19) (039.69, 429.19)      /T1_1 A	<|special_separator|>
-(044.62, 422.70) (272.88, 422.70) (272.88, 429.19) (044.62, 429.19)      /T1_2 )In-sampleresults.TheMongemodelistrainedasonemodelthatseesall	<|special_separator|>
-(306.47, 482.69) (555.73, 482.69) (555.73, 489.19) (306.47, 489.19)      /T1_2 conditionsbutisnotawareofconditionalinformation.TheCMongemodelsare	<|special_separator|>
-(306.47, 472.69) (546.94, 472.69) (546.94, 479.19) (306.47, 479.19)      /T1_2 likewisetrainedonallconditionsbutdogetconditionalinformationwiththe	<|special_separator|>
-(306.47, 462.69) (445.07, 462.69) (445.07, 469.19) (306.47, 469.19)      /T1_2 Mode-of-Action(MoA)orRDkitembedding.	<|special_separator|>
-(446.09, 462.83) (450.79, 462.83) (450.79, 469.19) (446.09, 469.19)      /T1_1 B	<|special_separator|>
-(450.82, 462.69) (545.92, 462.69) (545.92, 469.19) (450.82, 469.19)      /T1_2 )Out-of-sample(OOS)results.	<|special_separator|>
-(306.47, 452.69) (541.92, 452.69) (541.92, 459.18) (306.47, 459.18)      /T1_2 Allmodelsaretrainedinaleave-one-drug-outsetting.Thismeansthateach	<|special_separator|>
-(306.47, 442.68) (553.11, 442.68) (553.11, 449.18) (306.47, 449.18)      /T1_2 boxenplotshowstheresultover33drugs,from33models.TheMongemodelis	<|special_separator|>
-(306.47, 432.68) (559.20, 432.68) (559.20, 439.18) (306.47, 439.18)      /T1_2 alsotrainedinaleave-one-drug-outsettingbutdoesnotincorporateconditional	<|special_separator|>
-(306.47, 422.68) (347.11, 422.68) (347.11, 429.17) (306.47, 429.17)      /T1_2 information.

groundtruth/parser/font_11.pdf.page_no_18.py.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0ff4fefca209c5afa02afba0897dc7867af0d91815c70f7510e527cba88f2025
-size 1774411