Add files using upload-large-folder tool

.gitattributes

@@ -4519,3 +4519,239 @@ parse/train/BylQSxHFwr/BylQSxHFwr_span.pdf filter=lfs diff=lfs merge=lfs -text
 parse/train/HysBZSqlx/HysBZSqlx_origin.pdf filter=lfs diff=lfs merge=lfs -text
 parse/train/HysBZSqlx/HysBZSqlx_layout.pdf filter=lfs diff=lfs merge=lfs -text
 parse/train/HysBZSqlx/HysBZSqlx_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/KCd-3Pz8VjM/KCd-3Pz8VjM_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/KCd-3Pz8VjM/KCd-3Pz8VjM_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/KCd-3Pz8VjM/KCd-3Pz8VjM_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/LXMSvPmsm0g/LXMSvPmsm0g_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/LXMSvPmsm0g/LXMSvPmsm0g_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/LXMSvPmsm0g/LXMSvPmsm0g_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1l4eQW0Z/r1l4eQW0Z_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1l4eQW0Z/r1l4eQW0Z_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1l4eQW0Z/r1l4eQW0Z_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/JiYq3eqTKY/JiYq3eqTKY_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/JiYq3eqTKY/JiYq3eqTKY_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/JiYq3eqTKY/JiYq3eqTKY_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkEYojRqtm/SkEYojRqtm_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkEYojRqtm/SkEYojRqtm_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkEYojRqtm/SkEYojRqtm_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkxAS6VFDB/HkxAS6VFDB_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkxAS6VFDB/HkxAS6VFDB_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkxAS6VFDB/HkxAS6VFDB_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkIQH7qel/HkIQH7qel_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkIQH7qel/HkIQH7qel_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkIQH7qel/HkIQH7qel_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RUQ1zwZR8_/RUQ1zwZR8__origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RUQ1zwZR8_/RUQ1zwZR8__span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RUQ1zwZR8_/RUQ1zwZR8__layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJfWKsC5K7/SJfWKsC5K7_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJfWKsC5K7/SJfWKsC5K7_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJfWKsC5K7/SJfWKsC5K7_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/gV3wdEOGy_V/gV3wdEOGy_V_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/gV3wdEOGy_V/gV3wdEOGy_V_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/gV3wdEOGy_V/gV3wdEOGy_V_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJeHwJSYvH/SJeHwJSYvH_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJeHwJSYvH/SJeHwJSYvH_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJeHwJSYvH/SJeHwJSYvH_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJe4oxHYPB/BJe4oxHYPB_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJe4oxHYPB/BJe4oxHYPB_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJe4oxHYPB/BJe4oxHYPB_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1rRWl-Cb/H1rRWl-Cb_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1rRWl-Cb/H1rRWl-Cb_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1rRWl-Cb/H1rRWl-Cb_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJleph4KvS/BJleph4KvS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJleph4KvS/BJleph4KvS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJleph4KvS/BJleph4KvS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyfHgI6aW/HyfHgI6aW_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyfHgI6aW/HyfHgI6aW_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyfHgI6aW/HyfHgI6aW_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lJws05K7/H1lJws05K7_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lJws05K7/H1lJws05K7_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lJws05K7/H1lJws05K7_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B184E5qee/B184E5qee_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B184E5qee/B184E5qee_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B184E5qee/B184E5qee_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Rx9dBZaV_IP/Rx9dBZaV_IP_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Rx9dBZaV_IP/Rx9dBZaV_IP_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Rx9dBZaV_IP/Rx9dBZaV_IP_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByuI-mW0W/ByuI-mW0W_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByuI-mW0W/ByuI-mW0W_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByuI-mW0W/ByuI-mW0W_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RmuXDtjDhG/RmuXDtjDhG_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RmuXDtjDhG/RmuXDtjDhG_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/RmuXDtjDhG/RmuXDtjDhG_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJx0U64FwS/HJx0U64FwS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJx0U64FwS/HJx0U64FwS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJx0U64FwS/HJx0U64FwS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkVqXOxCb/SkVqXOxCb_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkVqXOxCb/SkVqXOxCb_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SkVqXOxCb/SkVqXOxCb_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S1dIzvclg/S1dIzvclg_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S1dIzvclg/S1dIzvclg_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S1dIzvclg/S1dIzvclg_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/N2dEPuK0vom/N2dEPuK0vom_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/N2dEPuK0vom/N2dEPuK0vom_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/N2dEPuK0vom/N2dEPuK0vom_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/qpsl2dR9twy/qpsl2dR9twy_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/qpsl2dR9twy/qpsl2dR9twy_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/qpsl2dR9twy/qpsl2dR9twy_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByfXe2C5tm/ByfXe2C5tm_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByfXe2C5tm/ByfXe2C5tm_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByfXe2C5tm/ByfXe2C5tm_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rkr1UDeC-/rkr1UDeC-_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rkr1UDeC-/rkr1UDeC-_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rkr1UDeC-/rkr1UDeC-_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/UoaQUQREMOs/UoaQUQREMOs_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/UoaQUQREMOs/UoaQUQREMOs_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/UoaQUQREMOs/UoaQUQREMOs_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/1ibNKMp8SKc/1ibNKMp8SKc_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/1ibNKMp8SKc/1ibNKMp8SKc_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/1ibNKMp8SKc/1ibNKMp8SKc_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BdKxQp0iBi8/BdKxQp0iBi8_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BdKxQp0iBi8/BdKxQp0iBi8_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BdKxQp0iBi8/BdKxQp0iBi8_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyDAQl-AW/HyDAQl-AW_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyDAQl-AW/HyDAQl-AW_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HyDAQl-AW/HyDAQl-AW_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-qh0M9XWxnv/-qh0M9XWxnv_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-qh0M9XWxnv/-qh0M9XWxnv_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-qh0M9XWxnv/-qh0M9XWxnv_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/nhkbYh30Tl/nhkbYh30Tl_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/nhkbYh30Tl/nhkbYh30Tl_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/nhkbYh30Tl/nhkbYh30Tl_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/CLnj31GZ4cI/CLnj31GZ4cI_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/CLnj31GZ4cI/CLnj31GZ4cI_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/CLnj31GZ4cI/CLnj31GZ4cI_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJeGZxrFvS/BJeGZxrFvS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJeGZxrFvS/BJeGZxrFvS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJeGZxrFvS/BJeGZxrFvS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Q1aiM7sCi1/Q1aiM7sCi1_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Q1aiM7sCi1/Q1aiM7sCi1_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Q1aiM7sCi1/Q1aiM7sCi1_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJypUGZ0Z/BJypUGZ0Z_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJypUGZ0Z/BJypUGZ0Z_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJypUGZ0Z/BJypUGZ0Z_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/C4-QQ1EHNcI/C4-QQ1EHNcI_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/C4-QQ1EHNcI/C4-QQ1EHNcI_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/C4-QQ1EHNcI/C4-QQ1EHNcI_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ryMxXPFex/ryMxXPFex_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ryMxXPFex/ryMxXPFex_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ryMxXPFex/ryMxXPFex_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1x-LjAcKX/B1x-LjAcKX_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1x-LjAcKX/B1x-LjAcKX_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1x-LjAcKX/B1x-LjAcKX_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ahYIlRBeCFw/ahYIlRBeCFw_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ahYIlRBeCFw/ahYIlRBeCFw_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ahYIlRBeCFw/ahYIlRBeCFw_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1E7Pwqgl/B1E7Pwqgl_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1E7Pwqgl/B1E7Pwqgl_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1E7Pwqgl/B1E7Pwqgl_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1eyO1BFPr/B1eyO1BFPr_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1eyO1BFPr/B1eyO1BFPr_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1eyO1BFPr/B1eyO1BFPr_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1ae1lZRb/B1ae1lZRb_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1ae1lZRb/B1ae1lZRb_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1ae1lZRb/B1ae1lZRb_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkz6aNqle/Hkz6aNqle_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkz6aNqle/Hkz6aNqle_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkz6aNqle/Hkz6aNqle_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyeD0RVtvS/SyeD0RVtvS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyeD0RVtvS/SyeD0RVtvS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyeD0RVtvS/SyeD0RVtvS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B16dGcqlx/B16dGcqlx_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B16dGcqlx/B16dGcqlx_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B16dGcqlx/B16dGcqlx_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/p2XgjS3Qp4X/p2XgjS3Qp4X_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/p2XgjS3Qp4X/p2XgjS3Qp4X_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/p2XgjS3Qp4X/p2XgjS3Qp4X_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/w3x8K0M6sAz/w3x8K0M6sAz_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/w3x8K0M6sAz/w3x8K0M6sAz_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/w3x8K0M6sAz/w3x8K0M6sAz_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkGTwjCctm/HkGTwjCctm_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkGTwjCctm/HkGTwjCctm_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkGTwjCctm/HkGTwjCctm_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hye_V0NKwr/Hye_V0NKwr_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hye_V0NKwr/Hye_V0NKwr_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hye_V0NKwr/Hye_V0NKwr_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rklwwo05Ym/rklwwo05Ym_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rklwwo05Ym/rklwwo05Ym_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rklwwo05Ym/rklwwo05Ym_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/eqBwg3AcIAK/eqBwg3AcIAK_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/eqBwg3AcIAK/eqBwg3AcIAK_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/eqBwg3AcIAK/eqBwg3AcIAK_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyJ7ClWCb/SyJ7ClWCb_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyJ7ClWCb/SyJ7ClWCb_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SyJ7ClWCb/SyJ7ClWCb_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Ozk9MrX1hvA/Ozk9MrX1hvA_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Ozk9MrX1hvA/Ozk9MrX1hvA_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Ozk9MrX1hvA/Ozk9MrX1hvA_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkcTe-bR-/HkcTe-bR-_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkcTe-bR-/HkcTe-bR-_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkcTe-bR-/HkcTe-bR-_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/D3PcGLdMx0/D3PcGLdMx0_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/D3PcGLdMx0/D3PcGLdMx0_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/D3PcGLdMx0/D3PcGLdMx0_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/cO1IH43yUF/cO1IH43yUF_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/cO1IH43yUF/cO1IH43yUF_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/cO1IH43yUF/cO1IH43yUF_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/St1giarCHLP/St1giarCHLP_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/St1giarCHLP/St1giarCHLP_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/St1giarCHLP/St1giarCHLP_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H12GRgcxg/H12GRgcxg_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H12GRgcxg/H12GRgcxg_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H12GRgcxg/H12GRgcxg_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HylUPnVKvH/HylUPnVKvH_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HylUPnVKvH/HylUPnVKvH_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HylUPnVKvH/HylUPnVKvH_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-msETI57gCH/-msETI57gCH_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-msETI57gCH/-msETI57gCH_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/-msETI57gCH/-msETI57gCH_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lBj2VFPS/H1lBj2VFPS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lBj2VFPS/H1lBj2VFPS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/H1lBj2VFPS/H1lBj2VFPS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S19eAF9ee/S19eAF9ee_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S19eAF9ee/S19eAF9ee_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/S19eAF9ee/S19eAF9ee_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJKkY35le/HJKkY35le_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJKkY35le/HJKkY35le_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HJKkY35le/HJKkY35le_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkg-xgrYvH/Hkg-xgrYvH_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkg-xgrYvH/Hkg-xgrYvH_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/Hkg-xgrYvH/Hkg-xgrYvH_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1YfAfcgl/B1YfAfcgl_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1YfAfcgl/B1YfAfcgl_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1YfAfcgl/B1YfAfcgl_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkwoSDPgg/HkwoSDPgg_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkwoSDPgg/HkwoSDPgg_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/HkwoSDPgg/HkwoSDPgg_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1hcZZ-AW/B1hcZZ-AW_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1hcZZ-AW/B1hcZZ-AW_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1hcZZ-AW/B1hcZZ-AW_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJfvknCqFQ/BJfvknCqFQ_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJfvknCqFQ/BJfvknCqFQ_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/BJfvknCqFQ/BJfvknCqFQ_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/n-219jrTht/n-219jrTht_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/n-219jrTht/n-219jrTht_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/n-219jrTht/n-219jrTht_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1QZ3zbAZ/r1QZ3zbAZ_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1QZ3zbAZ/r1QZ3zbAZ_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1QZ3zbAZ/r1QZ3zbAZ_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1laEnA5Ym/r1laEnA5Ym_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1laEnA5Ym/r1laEnA5Ym_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/r1laEnA5Ym/r1laEnA5Ym_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1l2bp4YwS/B1l2bp4YwS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1l2bp4YwS/B1l2bp4YwS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/B1l2bp4YwS/B1l2bp4YwS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJ1Xmf-Rb/SJ1Xmf-Rb_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJ1Xmf-Rb/SJ1Xmf-Rb_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/SJ1Xmf-Rb/SJ1Xmf-Rb_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/yhjpeuWepoj/yhjpeuWepoj_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/yhjpeuWepoj/yhjpeuWepoj_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/yhjpeuWepoj/yhjpeuWepoj_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByeL1R4FvS/ByeL1R4FvS_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByeL1R4FvS/ByeL1R4FvS_layout.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/ByeL1R4FvS/ByeL1R4FvS_origin.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rkfOvGbCW/rkfOvGbCW_span.pdf filter=lfs diff=lfs merge=lfs -text
+parse/train/rkfOvGbCW/rkfOvGbCW_origin.pdf filter=lfs diff=lfs merge=lfs -text

parse/dev/xT5rDp5VqKO/xT5rDp5VqKO.md

@@ -0,0 +1,134 @@
+# Coincidence Detection Is All You Need
+
+Anonymous Author(s)   
+Affiliation   
+Address   
+email
+
+# Abstract
+
+1 This paper demonstrates that the performance of coincidence detection - a classic   
+2 neuromorphic signal processing method found in Rosenblatt’s perceptrons with   
+3 distributed transmission times, can be competitive to a state-of-the-art deep learning   
+4 method for pattern recognition. Hence, we cannot remain comfortably numb to the   
+5 prevailing dogma that efficient matrix-vector operations is all we need; but should   
+6 enquire with greater vigour if more advanced continual learning methods (running   
+7 on spiking neural network hardware with neuromodulatory mechanisms at multiple   
+8 timescales) can beat the accuracy of task-specific deep learning methods.
+
+# 9 1 Introduction
+
+10 Frank Rosenblatt and his team (1957-1971) built and analyzed several kinds of perceptrons [1, 2, 3, 4]   
+11 - networks of sensory, association and receptor neurons; which in contemporary deep learning termi  
+12 nology relates to the input, hidden and output layers. The propagating signals were binary (compatible   
+13 with a spike-based view), the synaptic delays (transmission times) and weights (memory states) could   
+14 be analog, the network could be recurrent and was often randomly interconnected, and learning   
+15 often meant tuning the weights of the association-receptor subnetwork by some error-corrective   
+16 reinforcement. The synaptic delays were not learnt but instead randomly distributed in Rosenblatt’s   
+17 Tobermory perceptrons [5], and this was rich enough to realize concentration-invariant and uniform   
+18 time-warp invariant spatiotemporal classification by logarithmic encoding and coincidence detection.   
+19 However, the processing speed of commercial Von Neumann computers advanced exponentially   
+20 and outperformed neuromorphic hardware on yesterdecade’s benchmarks [6]. The Tobermory per  
+21 ceptron was forgotten, nevertheless, the utility of logarithmic encoding and coincidence detection   
+22 was formalized by John Hopfield [7] as an efficient solution to the analog match problem in pattern   
+23 recognition.   
+24 Now, half a century after the accidental demise of Rosenblatt, neuromorphic signal processors are   
+25 making a comeback. For example, (1) Intel’s Loihi with spike-time dependent plasticity mechanisms   
+26 for learning olfactory pattern recognizers [8]; (2) Physical reservoir computing networks [9] where   
+27 the interconnectivity of the hidden layer is unchanged, closer to the spirit of Rosenblatt’s randomly   
+28 interconnected sensory-association subnetwork.   
+29 Here, to strengthen the case for revisiting classic methods on novel and modern hardware, we evaluate   
+30 the performance of coincidence detection in comparison to a deep learning method. Nothing more,   
+31 nothing less, although this work was triggered by a rabid interest in employing artificial intelligence   
+32 to sniff out infections and prevent future pandemics.
+
+Table 1: Test accuracy $( \% )$   
+
+<table><tr><td>ResNet-26</td><td>Coincidence detection</td></tr><tr><td>82.2±0.3 (from [10])</td><td>82.7 (this work)</td></tr></table>
+
+# 33 2 Methods
+
+34 Here, we consider the work [10] of an interdisciplinary team, where a 26 layer convolutional neural   
+35 network with residual connections (ResNet-26) was successfully trained for classifying pathogenic   
+36 bacteria by Raman spectroscopy. In their work, there are $N = 3 0$ classes of bacterial isolates and   
+37 they begin with a ResNet-26 pre-trained on $N { \times } 2 0 0 0$ spectra, then for each class $n = 1 : N$ there are   
+38 $M = 1 0 0$ training spectra, and similarly $N \times M = 3 0 0 0$ test spectra. Each spectrum $_ { \textbf { \em x } }$ contains 1000   
+39 floating-point numbers ranging between 0 and 1. Although compute intensive, their deep learning   
+40 method proved to be a tool of great convenience for pattern recognition in a challenging dataset,   
+41 where intra-isolate spectra were often more dissimilar than inter-isolate spectra.   
+42 Our method to tackle the above dataset, is inspired by the theory of how coincidence detection [7]   
+43 in animal brains is fundamental for odour classification in complex and turbulent mixtures. Each   
+44 class $n$ has a vector representation ${ \pmb w } _ { n }$ that is learnt, and an input vector $_ { \textbf { \em x } }$ results in an output   
+45 class $y ( \pmb { x } ) = \arg _ { n } \operatorname* { m a x } ( \pmb { x } \wedge \pmb { w } _ { n } )$ where we introduce the operator $\Lambda$ to represent the coincidence   
+46 between two signals. The analytical nature of coincidence detection depends on the specificities of the   
+47 ion-channels and the membranes involved [11], and may even incorporate nonlinear leaky-integrate   
+48 [12] multiple timescale mechanisms. We do not yet have a complete theory of neuromorphic signal   
+49 processing, so here we introduce an approximation for the translation and scale-invariant property of   
+50 coincidence detection as
+
+$$
+\operatorname { a r g } _ { n } \operatorname * { m a x } ( { \pmb x } \bigwedge { \pmb w } _ { n } ) \approx \arg _ { n } \operatorname * { m a x } ( { \pmb w } _ { n } \cdot { \hat { \pmb x } } ) ,
+$$
+
+51 where $\hat { \pmb x }$ is the zero-mean unit-variance normalization of $_ { \textbf { \em x } }$ .
+
+52 Thus, the approximation in Eq. (1) allows $y ( \pmb { x } )$ to be learnt by a logistic regression on the normalized   
+53 dataset. We discard the pre-training data, pre-process the training and test spectra by a range-1 mean   
+54 filter, and use the default method for logistic regression in Wolfram Mathematica (L2-regularization   
+55 $= 0 . 0 0 0 1$ , optimization method $=$ limited-memory BFGS). Code is provided in the supplemental   
+56 material for reproducibility.
+
+# 57 3 Result and outlook
+
+58 The coincidence detection (via normalized logistic regression) method introduced here achieves a test   
+59 accuracy greater than ResNet-26 (see Table 1), and it took less than 3 seconds to train the classifier   
+60 on a modern desktop (without any special-purpose GPUs). Check the Appendix for a confusion   
+61 matrix plot of the training and test data. Note that the training data was fit all at once to a $100 \%$   
+62 accuracy. With a more neuromorphic coincidence detection method and a learning method that adapts   
+63 the synaptic delays $\pmb { w }$ continually, to keep track under changing environmental conditions, we may   
+64 achieve even greater accuracies.
+
+# 5 References
+
+6 [1] Frank Rosenblatt. The perceptron, a perceiving and recognizing automaton Project Para.   
+Cornell Aeronautical Laboratory, Inc. Report no. 85-460-1, 1957.   
+8 [2] Frank Rosenblatt. The perceptron: A theory of statistical separability in cognitive systems.   
+9 Cornell Aeronautical Laboratory, Inc. Report no. VG-1196-G-1, 1958.   
+0 [3] Frank Rosenblatt. Principles of neurodynamics. perceptrons and the theory of brain mechanisms.   
+1 Cornell Aeronautical Laboratory, Inc. Report no. 1196-G-8, 1961.
+
+[4] Frank Rosenblatt. Cognitive systems research program. Technical report, Cornell University, Ithaca, New York, 1971. [5] Frank Rosenblatt. A description of the tobermory perceptron. In Collected Technical Papers, volume 2. Cornell University, Ithaca, New York, 1963. [6] George Nagy. Neural networks-then and now. IEEE Transactions on Neural Networks, 2(2):316– 318, 1991.   
+[7] John J Hopfield. Pattern recognition computation using action potential timing for stimulus representation. Nature, 376(6535):33–36, 1995. [8] Nabil Imam and Thomas A Cleland. Rapid online learning and robust recall in a neuromorphic olfactory circuit. Nature Machine Intelligence, 2(3):181–191, 2020. [9] G. Tanaka, T. Yamane, J.B. Héroux, R. Nakane, N. Kanazawa, S. Takeda, H. Numata, D. Nakano, and A. Hirose. Recent advances in physical reservoir computing: A review. Neural Networks, 115:100–123, 2019.   
+[10] Chi-Sing Ho, Neal Jean, Catherine A Hogan, Lena Blackmon, Stefanie S Jeffrey, Mark Holodniy, Niaz Banaei, Amr AE Saleh, Stefano Ermon, and Jennifer Dionne. Rapid identification of pathogenic bacteria using raman spectroscopy and deep learning. Nature communications, 10(1):1–8, 2019.   
+[11] Nelson Spruston. Pyramidal neurons: dendritic structure and synaptic integration. Nature Reviews Neuroscience, 9(3):206–221, 2008.   
+[12] Wondimu Teka, Toma M Marinov, and Fidel Santamaria. Neuronal spike timing adaptation described with a fractional leaky integrate-and-fire model. PLoS computational biology, 10(3):e1003526, 2014.
+
+# 94 Checklist
+
+1. For all authors...
+
+(a) Do the main claims made in the abstract and introduction accurately reflect the paper’s contributions and scope? [Yes] See Table 1.   
+(b) Did you describe the limitations of your work? [Yes] Equation (1) makes it clear that we employ an approximation for coincidence detection.   
+(c) Did you discuss any potential negative societal impacts of your work? [N/A]   
+(d) Have you read the ethics review guidelines and ensured that your paper conforms to them? [Yes]
+
+2. If you are including theoretical results...
+
+(a) Did you state the full set of assumptions of all theoretical results? [N/A] (b) Did you include complete proofs of all theoretical results? [N/A]
+
+3. If you ran experiments...
+
+(a) Did you include the code, data, and instructions needed to reproduce the main experimental results (either in the supplemental material or as a URL)? [Yes] Check supplemental material   
+(b) Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes]   
+(c) Did you report error bars (e.g., with respect to the random seed after running experiments multiple times)? [N/A]   
+(d) Did you include the total amount of compute and the type of resources used (e.g., type of GPUs, internal cluster, or cloud provider)? [Yes] qualitatively, in the results section
+
+4. If you are using existing assets (e.g., code, data, models) or curating/releasing new assets...
+
+(a) If your work uses existing assets, did you cite the creators? [Yes]
+
+<table><tr><td>118</td><td>(b） Did you mention the license of the assets? [Yes] In the supplemental information</td></tr><tr><td>119</td><td>(c） Did you include any new assets either in the supplemental material or as a URL? [No]</td></tr><tr><td>120 121</td><td>(d) Did you discuss whether and how consent was obtained from people whose data you&#x27;re using/curating?[N/A]</td></tr><tr><td>122 123</td><td>(e） Did you discuss whether the data you are using/curating contains personally identifiable information or offensive content? [N/A]</td></tr><tr><td>124</td><td> 5. If you used crowdsourcing or conducted research with human subjects...</td></tr><tr><td>125 126</td><td>(a) Did you include the full text of instructions given to participants and screenshots, if applicable? [N/A]</td></tr><tr><td>127 128</td><td>(b） Did you describe any potential participant risks,with links to Institutional Review Board (IRB) approvals, if applicable?[N/A]</td></tr><tr><td>129 130</td><td>(c） Did you include the estimated hourly wage paid to participants and the total amount spent on participant compensation? [N/A]</td></tr></table>
+
+# 131 A Appendix
+
+![](images/5c386ac8dd4daa9b49b79785dc9443ed5b441416b3248608eedd54357405a51b.jpg)

parse/dev/xT5rDp5VqKO/xT5rDp5VqKO_content_list.json

@@ -0,0 +1,374 @@
+[
+    {
+        "type": "text",
+        "text": "Coincidence Detection Is All You Need ",
+        "text_level": 1,
+        "bbox": [
+            264,
+            122,
+            732,
+            146
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "Anonymous Author(s)   \nAffiliation   \nAddress   \nemail ",
+        "bbox": [
+            423,
+            196,
+            580,
+            252
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "Abstract ",
+        "text_level": 1,
+        "bbox": [
+            462,
+            287,
+            535,
+            304
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "1 This paper demonstrates that the performance of coincidence detection - a classic   \n2 neuromorphic signal processing method found in Rosenblatt’s perceptrons with   \n3 distributed transmission times, can be competitive to a state-of-the-art deep learning   \n4 method for pattern recognition. Hence, we cannot remain comfortably numb to the   \n5 prevailing dogma that efficient matrix-vector operations is all we need; but should   \n6 enquire with greater vigour if more advanced continual learning methods (running   \n7 on spiking neural network hardware with neuromodulatory mechanisms at multiple   \n8 timescales) can beat the accuracy of task-specific deep learning methods. ",
+        "bbox": [
+            150,
+            329,
+            766,
+            440
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "9 1 Introduction ",
+        "text_level": 1,
+        "bbox": [
+            151,
+            512,
+            338,
+            532
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "10 Frank Rosenblatt and his team (1957-1971) built and analyzed several kinds of perceptrons [1, 2, 3, 4]   \n11 - networks of sensory, association and receptor neurons; which in contemporary deep learning termi  \n12 nology relates to the input, hidden and output layers. The propagating signals were binary (compatible   \n13 with a spike-based view), the synaptic delays (transmission times) and weights (memory states) could   \n14 be analog, the network could be recurrent and was often randomly interconnected, and learning   \n15 often meant tuning the weights of the association-receptor subnetwork by some error-corrective   \n16 reinforcement. The synaptic delays were not learnt but instead randomly distributed in Rosenblatt’s   \n17 Tobermory perceptrons [5], and this was rich enough to realize concentration-invariant and uniform   \n18 time-warp invariant spatiotemporal classification by logarithmic encoding and coincidence detection.   \n19 However, the processing speed of commercial Von Neumann computers advanced exponentially   \n20 and outperformed neuromorphic hardware on yesterdecade’s benchmarks [6]. The Tobermory per  \n21 ceptron was forgotten, nevertheless, the utility of logarithmic encoding and coincidence detection   \n22 was formalized by John Hopfield [7] as an efficient solution to the analog match problem in pattern   \n23 recognition.   \n24 Now, half a century after the accidental demise of Rosenblatt, neuromorphic signal processors are   \n25 making a comeback. For example, (1) Intel’s Loihi with spike-time dependent plasticity mechanisms   \n26 for learning olfactory pattern recognizers [8]; (2) Physical reservoir computing networks [9] where   \n27 the interconnectivity of the hidden layer is unchanged, closer to the spirit of Rosenblatt’s randomly   \n28 interconnected sensory-association subnetwork.   \n29 Here, to strengthen the case for revisiting classic methods on novel and modern hardware, we evaluate   \n30 the performance of coincidence detection in comparison to a deep learning method. Nothing more,   \n31 nothing less, although this work was triggered by a rabid interest in employing artificial intelligence   \n32 to sniff out infections and prevent future pandemics. ",
+        "bbox": [
+            147,
+            558,
+            825,
+            750
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "",
+        "bbox": [
+            147,
+            756,
+            825,
+            825
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "text",
+        "text": "",
+        "bbox": [
+            147,
+            832,
+            825,
+            888
+        ],
+        "page_idx": 0
+    },
+    {
+        "type": "table",
+        "img_path": "images/652089d176df3cac32600cfc9074e1ab496373b1c6181602a93c29960ad2ed13.jpg",
+        "table_caption": [
+            "Table 1: Test accuracy $( \\% )$ "
+        ],
+        "table_footnote": [],
+        "table_body": "<table><tr><td>ResNet-26</td><td>Coincidence detection</td></tr><tr><td>82.2±0.3 (from [10])</td><td>82.7 (this work)</td></tr></table>",
+        "bbox": [
+            333,
+            109,
+            663,
+            155
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "33 2 Methods ",
+        "text_level": 1,
+        "bbox": [
+            147,
+            175,
+            299,
+            195
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "34 Here, we consider the work [10] of an interdisciplinary team, where a 26 layer convolutional neural   \n35 network with residual connections (ResNet-26) was successfully trained for classifying pathogenic   \n36 bacteria by Raman spectroscopy. In their work, there are $N = 3 0$ classes of bacterial isolates and   \n37 they begin with a ResNet-26 pre-trained on $N { \\times } 2 0 0 0$ spectra, then for each class $n = 1 : N$ there are   \n38 $M = 1 0 0$ training spectra, and similarly $N \\times M = 3 0 0 0$ test spectra. Each spectrum $_ { \\textbf { \\em x } }$ contains 1000   \n39 floating-point numbers ranging between 0 and 1. Although compute intensive, their deep learning   \n40 method proved to be a tool of great convenience for pattern recognition in a challenging dataset,   \n41 where intra-isolate spectra were often more dissimilar than inter-isolate spectra.   \n42 Our method to tackle the above dataset, is inspired by the theory of how coincidence detection [7]   \n43 in animal brains is fundamental for odour classification in complex and turbulent mixtures. Each   \n44 class $n$ has a vector representation ${ \\pmb w } _ { n }$ that is learnt, and an input vector $_ { \\textbf { \\em x } }$ results in an output   \n45 class $y ( \\pmb { x } ) = \\arg _ { n } \\operatorname* { m a x } ( \\pmb { x } \\wedge \\pmb { w } _ { n } )$ where we introduce the operator $\\Lambda$ to represent the coincidence   \n46 between two signals. The analytical nature of coincidence detection depends on the specificities of the   \n47 ion-channels and the membranes involved [11], and may even incorporate nonlinear leaky-integrate   \n48 [12] multiple timescale mechanisms. We do not yet have a complete theory of neuromorphic signal   \n49 processing, so here we introduce an approximation for the translation and scale-invariant property of   \n50 coincidence detection as ",
+        "bbox": [
+            145,
+            213,
+            825,
+            325
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "",
+        "bbox": [
+            147,
+            330,
+            825,
+            455
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "equation",
+        "img_path": "images/47034e52b12da81112f808e1a0b9eb8beeb7452f9b1bf3394fcb5a7e623378a3.jpg",
+        "text": "$$\n\\operatorname { a r g } _ { n } \\operatorname * { m a x } ( { \\pmb x } \\bigwedge { \\pmb w } _ { n } ) \\approx \\arg _ { n } \\operatorname * { m a x } ( { \\pmb w } _ { n } \\cdot { \\hat { \\pmb x } } ) ,\n$$",
+        "text_format": "latex",
+        "bbox": [
+            354,
+            462,
+            642,
+            483
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "51 where $\\hat { \\pmb x }$ is the zero-mean unit-variance normalization of $_ { \\textbf { \\em x } }$ . ",
+        "bbox": [
+            147,
+            491,
+            563,
+            506
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "52 Thus, the approximation in Eq. (1) allows $y ( \\pmb { x } )$ to be learnt by a logistic regression on the normalized   \n53 dataset. We discard the pre-training data, pre-process the training and test spectra by a range-1 mean   \n54 filter, and use the default method for logistic regression in Wolfram Mathematica (L2-regularization   \n55 $= 0 . 0 0 0 1$ , optimization method $=$ limited-memory BFGS). Code is provided in the supplemental   \n56 material for reproducibility. ",
+        "bbox": [
+            147,
+            511,
+            825,
+            582
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "57 3 Result and outlook ",
+        "text_level": 1,
+        "bbox": [
+            150,
+            609,
+            400,
+            630
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "58 The coincidence detection (via normalized logistic regression) method introduced here achieves a test   \n59 accuracy greater than ResNet-26 (see Table 1), and it took less than 3 seconds to train the classifier   \n60 on a modern desktop (without any special-purpose GPUs). Check the Appendix for a confusion   \n61 matrix plot of the training and test data. Note that the training data was fit all at once to a $100 \\%$   \n62 accuracy. With a more neuromorphic coincidence detection method and a learning method that adapts   \n63 the synaptic delays $\\pmb { w }$ continually, to keep track under changing environmental conditions, we may   \n64 achieve even greater accuracies. ",
+        "bbox": [
+            147,
+            648,
+            825,
+            746
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "5 References ",
+        "text_level": 1,
+        "bbox": [
+            156,
+            775,
+            284,
+            794
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "6 [1] Frank Rosenblatt. The perceptron, a perceiving and recognizing automaton Project Para.   \nCornell Aeronautical Laboratory, Inc. Report no. 85-460-1, 1957.   \n8 [2] Frank Rosenblatt. The perceptron: A theory of statistical separability in cognitive systems.   \n9 Cornell Aeronautical Laboratory, Inc. Report no. VG-1196-G-1, 1958.   \n0 [3] Frank Rosenblatt. Principles of neurodynamics. perceptrons and the theory of brain mechanisms.   \n1 Cornell Aeronautical Laboratory, Inc. Report no. 1196-G-8, 1961. ",
+        "bbox": [
+            156,
+            806,
+            830,
+            912
+        ],
+        "page_idx": 1
+    },
+    {
+        "type": "text",
+        "text": "[4] Frank Rosenblatt. Cognitive systems research program. Technical report, Cornell University, Ithaca, New York, 1971. [5] Frank Rosenblatt. A description of the tobermory perceptron. In Collected Technical Papers, volume 2. Cornell University, Ithaca, New York, 1963. [6] George Nagy. Neural networks-then and now. IEEE Transactions on Neural Networks, 2(2):316– 318, 1991.   \n[7] John J Hopfield. Pattern recognition computation using action potential timing for stimulus representation. Nature, 376(6535):33–36, 1995. [8] Nabil Imam and Thomas A Cleland. Rapid online learning and robust recall in a neuromorphic olfactory circuit. Nature Machine Intelligence, 2(3):181–191, 2020. [9] G. Tanaka, T. Yamane, J.B. Héroux, R. Nakane, N. Kanazawa, S. Takeda, H. Numata, D. Nakano, and A. Hirose. Recent advances in physical reservoir computing: A review. Neural Networks, 115:100–123, 2019.   \n[10] Chi-Sing Ho, Neal Jean, Catherine A Hogan, Lena Blackmon, Stefanie S Jeffrey, Mark Holodniy, Niaz Banaei, Amr AE Saleh, Stefano Ermon, and Jennifer Dionne. Rapid identification of pathogenic bacteria using raman spectroscopy and deep learning. Nature communications, 10(1):1–8, 2019.   \n[11] Nelson Spruston. Pyramidal neurons: dendritic structure and synaptic integration. Nature Reviews Neuroscience, 9(3):206–221, 2008.   \n[12] Wondimu Teka, Toma M Marinov, and Fidel Santamaria. Neuronal spike timing adaptation described with a fractional leaky integrate-and-fire model. PLoS computational biology, 10(3):e1003526, 2014. ",
+        "bbox": [
+            158,
+            93,
+            828,
+            477
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "94 Checklist ",
+        "text_level": 1,
+        "bbox": [
+            148,
+            508,
+            269,
+            527
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "1. For all authors... ",
+        "bbox": [
+            214,
+            542,
+            339,
+            556
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "(a) Do the main claims made in the abstract and introduction accurately reflect the paper’s contributions and scope? [Yes] See Table 1.   \n(b) Did you describe the limitations of your work? [Yes] Equation (1) makes it clear that we employ an approximation for coincidence detection.   \n(c) Did you discuss any potential negative societal impacts of your work? [N/A]   \n(d) Have you read the ethics review guidelines and ensured that your paper conforms to them? [Yes] ",
+        "bbox": [
+            238,
+            560,
+            825,
+            666
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "2. If you are including theoretical results... ",
+        "bbox": [
+            214,
+            670,
+            493,
+            684
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "(a) Did you state the full set of assumptions of all theoretical results? [N/A] (b) Did you include complete proofs of all theoretical results? [N/A] ",
+        "bbox": [
+            238,
+            688,
+            736,
+            719
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "3. If you ran experiments... ",
+        "bbox": [
+            214,
+            724,
+            393,
+            738
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "(a) Did you include the code, data, and instructions needed to reproduce the main experimental results (either in the supplemental material or as a URL)? [Yes] Check supplemental material   \n(b) Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes]   \n(c) Did you report error bars (e.g., with respect to the random seed after running experiments multiple times)? [N/A]   \n(d) Did you include the total amount of compute and the type of resources used (e.g., type of GPUs, internal cluster, or cloud provider)? [Yes] qualitatively, in the results section ",
+        "bbox": [
+            238,
+            742,
+            825,
+            875
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "4. If you are using existing assets (e.g., code, data, models) or curating/releasing new assets... ",
+        "bbox": [
+            220,
+            878,
+            825,
+            893
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "text",
+        "text": "(a) If your work uses existing assets, did you cite the creators? [Yes] ",
+        "bbox": [
+            235,
+            897,
+            694,
+            911
+        ],
+        "page_idx": 2
+    },
+    {
+        "type": "table",
+        "img_path": "images/bbde7388781d76919dd279438b8326026bd9d7e50f61bcf96f594a7c30ade3ff.jpg",
+        "table_caption": [],
+        "table_footnote": [],
+        "table_body": "<table><tr><td>118</td><td>(b） Did you mention the license of the assets? [Yes] In the supplemental information</td></tr><tr><td>119</td><td>(c） Did you include any new assets either in the supplemental material or as a URL? [No]</td></tr><tr><td>120 121</td><td>(d) Did you discuss whether and how consent was obtained from people whose data you&#x27;re using/curating?[N/A]</td></tr><tr><td>122 123</td><td>(e） Did you discuss whether the data you are using/curating contains personally identifiable information or offensive content? [N/A]</td></tr><tr><td>124</td><td> 5. If you used crowdsourcing or conducted research with human subjects...</td></tr><tr><td>125 126</td><td>(a) Did you include the full text of instructions given to participants and screenshots, if applicable? [N/A]</td></tr><tr><td>127 128</td><td>(b） Did you describe any potential participant risks,with links to Institutional Review Board (IRB) approvals, if applicable?[N/A]</td></tr><tr><td>129 130</td><td>(c） Did you include the estimated hourly wage paid to participants and the total amount spent on participant compensation? [N/A]</td></tr></table>",
+        "bbox": [
+            133,
+            71,
+            830,
+            647
+        ],
+        "page_idx": 3
+    },
+    {
+        "type": "text",
+        "text": "131 A Appendix ",
+        "text_level": 1,
+        "bbox": [
+            142,
+            813,
+            313,
+            833
+        ],
+        "page_idx": 3
+    },
+    {
+        "type": "image",
+        "img_path": "images/5c386ac8dd4daa9b49b79785dc9443ed5b441416b3248608eedd54357405a51b.jpg",
+        "image_caption": [],
+        "image_footnote": [],
+        "bbox": [
+            179,
+            74,
+            709,
+            915
+        ],
+        "page_idx": 4
+    }
+]

parse/train/-msETI57gCH/-msETI57gCH_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5ef554da5b5f24990e3cc30ead4e9f8c97251ca834954986850a6c68f6e102a
+size 771648

parse/train/-msETI57gCH/-msETI57gCH_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dde1fa5aac63bf2eca3fbf668bbf8777da17626c9e9cfb02bfcaf6c003f97f0c
+size 538010

parse/train/-msETI57gCH/-msETI57gCH_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94eec0c5ea2ab09e5e97da01d537f3747c7cc74ab760d1718876f8ead51e47c2
+size 797873

parse/train/-qh0M9XWxnv/-qh0M9XWxnv_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0347700c6d5388f1ea4ee2b2711c8c373055b116b77111a7def8224afd78f23
+size 3659250

parse/train/-qh0M9XWxnv/-qh0M9XWxnv_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eda9fbb7e7e3506b8b88aca5c1b8abc803f1c550b233277ecd8e402bcaf169aa
+size 3358105

parse/train/-qh0M9XWxnv/-qh0M9XWxnv_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b93727dcb885753c8b127afb2c754059176a9dafff9669f4a762d87d79c59443
+size 3665278

parse/train/1ibNKMp8SKc/1ibNKMp8SKc_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28da3c6c5426476b8839c751742cbf20b3dd4907ab195767a0598628f9b6bcdd
+size 12752305

parse/train/1ibNKMp8SKc/1ibNKMp8SKc_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e547e7cf09bde698be36b04dc291f29afe9490e698e2dd0c0d12d646dcfb6ac
+size 12592959

parse/train/1ibNKMp8SKc/1ibNKMp8SKc_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71e88613533c5d1954fa9beaf629d6e8890118ab840b6a274ffe615211e62930
+size 12754707

parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b924890d4e4ae1f0e626a686096f1a76f3cd16e550710daa77ef2fcab2d69c35
+size 517131

parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9225d90db7b40b1b1e043a409477e3796754d0a9d08d988ef8d4ed7720154e4
+size 350396

parse/train/8PPmLbV6WiQ/8PPmLbV6WiQ_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2342bee043c4b9ab56dfede8e465ff05ae04ca04c20fb7a80440f399d35e2d8f
+size 543909

parse/train/B16dGcqlx/B16dGcqlx_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85425c52ff8d1657933127e247029ae7c82e905ef7c50445d85ab09157a96fe9
+size 2196108

parse/train/B16dGcqlx/B16dGcqlx_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0673f681dec9ba6ddf86a6e2b5c0a724e8dcc6e92d757d09af77330904655c0
+size 2070289

parse/train/B16dGcqlx/B16dGcqlx_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14075703ace3c48d4ff4ba07ef5d284521cf3d73dbad3cc1fdb1b54e2015ab27
+size 2198877

parse/train/B184E5qee/B184E5qee_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c640e72b04e4622c922fad6445baf4f88a4bff2f7cf9d656e558ec51cfbdfc50
+size 716267

parse/train/B184E5qee/B184E5qee_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d119483629d2c40f7c888201c4b50b40b97dcc8a69a3e43654344cab6a6f4c6
+size 615534

parse/train/B184E5qee/B184E5qee_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:670a3969080c64399c51a1d82bb3376dd157520af7bc106748e2e5304302468d
+size 713949

parse/train/B1E7Pwqgl/B1E7Pwqgl_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac5af678584e3a169a20ca2d843f5715613d13bbdb33033f9151051edc1dd086
+size 64646217

parse/train/B1E7Pwqgl/B1E7Pwqgl_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1337ac8a32b4aea9038816bf29069a73cd354092d9b11ae9536ff0c732016f94
+size 64453240

parse/train/B1E7Pwqgl/B1E7Pwqgl_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75eebcf569999c2acd539a695d8fde2d27096118b6a14c59d1f33cf1927f6f72
+size 64648455

parse/train/B1YfAfcgl/B1YfAfcgl_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0230963c38218156b0d5a0540b37e63e4ac7738b2d362c204f4799b923e780fe
+size 899744

parse/train/B1YfAfcgl/B1YfAfcgl_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db367f0087fe63b39a3566d823aad653387ab9363d8497e79f5c086bba5baa0d
+size 673440

parse/train/B1YfAfcgl/B1YfAfcgl_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:149b9eec14e9a52254b9d956fd8b661c9ba5a7587c832309a3677dde26bb5941
+size 912967

parse/train/B1ae1lZRb/B1ae1lZRb_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:873eedd9aeda0550f66e38bfce6d962ac0ab9544a59aa6edc14b7b0866befd3c
+size 575882

parse/train/B1ae1lZRb/B1ae1lZRb_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7914f66825e160779bc67fafd9513c591ef7a83ec8b87d3edef1d3a546e64f2d
+size 439027

parse/train/B1ae1lZRb/B1ae1lZRb_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c844449960aa406d1149584d24b670a4ca783ebdb41daeb7654b583878a8e09d
+size 582274

parse/train/B1eyO1BFPr/B1eyO1BFPr_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4493a29ba0d43689589dc93c5834f77e4e5c78c992f34e3da0c60b3a175581af
+size 4585539

parse/train/B1eyO1BFPr/B1eyO1BFPr_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a3faf4c91601378cde0bebc914098f7bc6f8cc0268dac4b6a602a592034ec6c
+size 4035107

parse/train/B1eyO1BFPr/B1eyO1BFPr_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cde72629570ecb98103ed3d3af0569d0f1d34ce9fc4a701d9c2988673760d98
+size 4606455

parse/train/B1hcZZ-AW/B1hcZZ-AW_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:738b0541b40ea31b7057e9db65815b8ef4ee853b60844af089417cc51d75ed10
+size 9005394

parse/train/B1hcZZ-AW/B1hcZZ-AW_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4eb440cd70d1cfd1e4360b3b369c0492ee726176de5b6da7daf0f0aa5d566ba
+size 8838086

parse/train/B1hcZZ-AW/B1hcZZ-AW_span.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eba803ea572705eb22434f2c590bfee0fb2c263aeefc4d589c0bf4df12b7ac90
+size 9000543

parse/train/B1l2bp4YwS/B1l2bp4YwS_layout.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cef2394ce519b8a299d7689d9dcb1ebbbd58d8bb004c5844aef3a7680818c4a
+size 720442

parse/train/B1l2bp4YwS/B1l2bp4YwS_origin.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3769e1b527cedca3be0be02bb9616ca9ca6721efd730829056c5fd78a139b6b7
+size 502709