Datasets:
Tasks:
Question Answering
Modalities:
Text
Formats:
json
Languages:
English
Size:
1K - 10K
Tags:
document-question-answering
visual-question-answering
multimodal
grounding
scientific-documents
query string | answer string | doc_name string | evidence_page list | bbox list | rel_bbox list | subimg_type list | category string |
|---|---|---|---|---|---|---|---|
In the figure, which letter uniquely receives two arrows, and from which letters do they come? | D; from C and E. | 2311.07631 | [
4
] | [
[
[
281,
2228,
1890,
2941
]
]
] | [
[
[
110.196078,
675.151515,
741.176471,
891.212121
]
]
] | [
[
"image"
]
] | cs |
In the identities table, what AI entry corresponds to the DS term 'rational'? | machine | 2311.07631 | [
26
] | [
[
[
305,
1512,
1520,
2005
]
]
] | [
[
[
119.607843,
458.181818,
596.078431,
607.575758
]
]
] | [
[
"table"
]
] | cs |
For rows with focus 'human', list each area and its first-listed topic in table order. | design—ontologies; value—ethics | 2311.07631 | [
26
] | [
[
[
287,
2025,
2263,
2527
]
]
] | [
[
[
112.54902,
613.636364,
887.45098,
765.757576
]
]
] | [
[
"table"
]
] | cs |
Which AREA numerals show a prime mark, and how many named DOMAINS (excluding 'etc') are listed? | II' and III'; four. | 2311.07631 | [
8
] | [
[
[
546,
2096,
1987,
2812
]
]
] | [
[
[
214.117647,
635.151515,
779.215686,
852.121212
]
]
] | [
[
"image"
]
] | cs |
Listing the right-hand terms from top to bottom in DS then AI, what sequence appears? | concrete, machine, action, rational | 2311.07631 | [
26
] | [
[
[
874,
1238,
1708,
1494
]
]
] | [
[
[
342.745098,
375.151515,
669.803922,
452.727273
]
]
] | [
[
"table"
]
] | cs |
In Table 1, which two cells contain primed Roman numerals, and what keywords appear beneath them? | Human-Abstract DESIGN; Machine-Concrete SYSTEMS | 2311.07631 | [
9
] | [
[
[
719,
1244,
1796,
1770
]
]
] | [
[
[
281.960784,
376.969697,
704.313725,
536.363636
]
]
] | [
[
"table"
]
] | cs |
In the described spiral process, between which two poles does it oscillate, and what page number is cited for the processual definition? | Diversification and simplification; p. 41. | 2311.07631 | [
15
] | [
[
[
290,
2025,
2245,
2985
]
]
] | [
[
[
113.72549,
613.636364,
880.392157,
904.545455
]
]
] | [
[
"text"
]
] | cs |
In the excerpt, what verb describes how the two principal components relate to the general field of data science? | undergird | 2311.07631 | [
9
] | [
[
[
293,
2812,
2187,
2991
]
]
] | [
[
[
114.901961,
852.121212,
857.647059,
906.363636
]
]
] | [
[
"text"
]
] | cs |
In the passage, which area contains people good at working with abstract machinery, and what does the model describe instead of objective reality? | The analytics area; areas of academic expertise. | 2311.07631 | [
11
] | [
[
[
293,
2436,
2222,
2888
]
]
] | [
[
[
114.901961,
738.181818,
871.372549,
875.151515
]
]
] | [
[
"text"
]
] | cs |
According to the sentence, what two claims are made about definitions' inclusiveness and relative breadth? | None includes them all; some are more comprehensive. | 2311.07631 | [
3
] | [
[
[
287,
1826,
2196,
1937
]
]
] | [
[
[
112.54902,
553.333333,
861.176471,
586.969697
]
]
] | [
[
"text"
]
] | cs |
In Figure 1’s confusion matrix, which utility label and exact probability symbol correspond to the true positives (positive case)? | U11 and p^s_11 | 2405.09360 | [
2
] | [
[
[
1324,
249,
2351,
1074
]
]
] | [
[
[
519.215686,
75.454545,
921.960784,
325.454545
]
]
] | [
[
"image"
]
] | econ |
In the proof’s construction with PD(Y=1, Ŷ=1)>0, which parameter equalities set UD3 and UD2 to zero, and what value is chosen for U10? | U01=U10; U00=U10; U10=0. | 2405.09360 | [
11
] | [
[
[
211,
1987,
1241,
2935
]
]
] | [
[
[
82.745098,
602.121212,
486.666667,
889.393939
]
]
] | [
[
"image"
]
] | econ |
In the 'Mortgage example' plot, what is the numerical difference between the mean μ of X_T and the barrier b shown in the legend? | 20,000 | 2405.09360 | [
8
] | [
[
[
1309,
205,
2339,
998
]
]
] | [
[
[
513.333333,
62.121212,
917.254902,
302.424242
]
]
] | [
[
"image"
]
] | econ |
Under Proposition 12 (U ≥ 0), the upper bound of the utility difference is given by what product, and which equation number labels it? | tilde U times Π*, equation (20) | 2405.09360 | [
8
] | [
[
[
214,
1262,
1227,
1650
]
]
] | [
[
[
83.921569,
382.424242,
481.176471,
500
]
]
] | [
[
"image"
]
] | econ |
In Figure 3's confusion matrix, which U-region spans both Y=0 and Y=1 under Ŷ=0, and what gray p^s term appears inside it? | U_0; p^s_0 | 2405.09360 | [
6
] | [
[
[
1327,
252,
2334,
989
]
]
] | [
[
[
520.392157,
76.363636,
915.294118,
299.69697
]
]
] | [
[
"image"
]
] | econ |
Under the text’s fairness notion at level τ = 0, what inequality must the utility difference UD satisfy? | UD ≤ 0 | 2405.09360 | [
3
] | [
[
[
214,
918,
1218,
1814
]
]
] | [
[
[
83.921569,
278.181818,
477.647059,
549.69697
]
]
] | [
[
"image"
]
] | econ |
In the excerpt, what is the equation number for Es[U] = U·Ps + U10, and which probabilities must coincide for fairness to hold? | 5; joint probabilities of both groups. | 2405.09360 | [
3
] | [
[
[
1312,
223,
2337,
601
]
]
] | [
[
[
514.509804,
67.575758,
916.470588,
182.121212
]
]
] | [
[
"image"
]
] | econ |
According to equation (18), (C)UA holds if which two PD expressions are zero? | PD(Y=1|Ŷ=1) and PD(Ŷ=0) | 2405.09360 | [
6
] | [
[
[
1327,
2322,
2331,
2827
]
]
] | [
[
[
520.392157,
703.636364,
914.117647,
856.666667
]
]
] | [
[
"image"
]
] | econ |
In Proposition 8’s second claim, what factor multiplies ε·K in the no-disadvantage bound, and which norm of PD is used in the inequality for |UD|? | 1 + 2Γ; the infinity norm ||PD||∞. | 2405.09360 | [
11
] | [
[
[
1324,
731,
2339,
2938
]
]
] | [
[
[
519.215686,
221.515152,
917.254902,
890.30303
]
]
] | [
[
"image"
]
] | econ |
In Proposition 7, assuming statistical parity and equalized odds with U_01 = U_10, what value is U_3 concluded to have, and which equation number is cited for the claim? | U_3 = 0; Equation (7). | 2405.09360 | [
11
] | [
[
[
1315,
231,
2328,
707
]
]
] | [
[
[
515.686275,
70,
912.941176,
214.242424
]
]
] | [
[
"image"
]
] | econ |
According to Theorem 1, under what tail relation does asymptotic learning of informativeness fail? | Uninformative signals have thinner tails than informative signals. | 2406.05299 | [
9
] | [
[
[
290,
1161,
2200,
1307
]
]
] | [
[
[
116.888351,
330.957811,
886.739218,
372.576967
]
]
] | [
[
"image"
]
] | econ |
Conditioned on θ and ω=1, when a_t = b, what are both the exact update for r_{t+1} and the probability of that action (include signs)? | r_{t+1}=r_t+D_b(r_t); probability F_θ(-r_t). | 2406.05299 | [
10
] | [
[
[
290,
2131,
2190,
2955
]
]
] | [
[
[
116.888351,
607.468643,
882.708585,
842.360319
]
]
] | [
[
"image"
]
] | econ |
In Proposition 2, conditioning on which ω value must both events {ā = g} and {ā = b} have zero probability for asymptotic learning of informativeness to hold? | ω = 0 | 2406.05299 | [
12
] | [
[
[
287,
905,
2187,
1048
]
]
] | [
[
[
115.679162,
257.981756,
881.499395,
298.745724
]
]
] | [
[
"image"
]
] | econ |
In Claim 2’s displayed equality, when conditioning on π_t = π, which probability measure appears on the right-hand side? | P_{tilde ω, tilde θ, π} | 2406.05299 | [
11
] | [
[
[
284,
736,
2194,
1376
]
]
] | [
[
[
114.469972,
209.806157,
884.320838,
392.246294
]
]
] | [
[
"image"
]
] | econ |
In the tail-thickness definitions, which specific L_θ and R_θ are bounded below by ε for fatter tails, and which are bounded above by 1/ε for thinner tails? | L_b and R_g ≥ ε; L_g or R_b ≤ 1/ε. | 2406.05299 | [
8
] | [
[
[
280,
1239,
2187,
2490
]
]
] | [
[
[
112.857719,
353.192702,
881.499395,
709.806157
]
]
] | [
[
"image"
]
] | econ |
In equation (A.1), what term multiplies P[ω = 0]? | P0[A∞(k)] | 2406.05299 | [
16
] | [
[
[
293,
699,
2184,
1888
]
]
] | [
[
[
118.097541,
199.258837,
880.290206,
538.198404
]
]
] | [
[
"image"
]
] | econ |
In the proof, what explicit expression does Bayes’ rule give for the ratio P0[q_t = q̃] divided by P1[q_t = q̃]? | (1−q̃)/q̃ × γ/(1−γ) | 2406.05299 | [
15
] | [
[
[
280,
1860,
2200,
3011
]
]
] | [
[
[
112.857719,
530.216648,
886.739218,
858.323831
]
]
] | [
[
"image"
]
] | econ |
In the log-odds formulations, what variables are conditioned on for r_t versus L_t? | r_t: H_t; L_t: H_t and s_t. | 2406.05299 | [
10
] | [
[
[
290,
1154,
2190,
2081
]
]
] | [
[
[
116.888351,
328.962372,
882.708585,
593.215507
]
]
] | [
[
"image"
]
] | econ |
Using the symmetry r_t^b = −r_t^g, in which numbered equation does the condition become ∑_{t=1}^∞ (1 − F_g(•)) = ∞, and what is •? | (6); r_t^g | 2406.05299 | [
12
] | [
[
[
290,
1391,
2200,
3158
]
]
] | [
[
[
116.888351,
396.522235,
886.739218,
900.22805
]
]
] | [
[
"image"
]
] | econ |
In the proof, what principle gives the bound on π_{τ+1}, and what property justifies the final equality in the chain of expectations? | Overturning principle; stationarity property. | 2406.05299 | [
17
] | [
[
[
284,
1257,
2187,
2428
]
]
] | [
[
[
114.469972,
358.323831,
881.499395,
692.132269
]
]
] | [
[
"image"
]
] | econ |
What exact equation characterizes α* where the randomized payoff is maximized? | α* + ln(1−α*)·(1−α*) = c0 | 2406.11528 | [
11
] | [
[
[
302,
584,
2193,
1773
]
]
] | [
[
[
118.431373,
176.969697,
860,
537.272727
]
]
] | [
[
"image"
]
] | econ |
According to the proof, after rewriting the limit in x, what exact expression appears under the square root in the denominator? | x + ln(1 − x) · (1 − x) | 2406.11528 | [
11
] | [
[
[
296,
1808,
2257,
2486
]
]
] | [
[
[
116.078431,
547.878788,
885.098039,
753.333333
]
]
] | [
[
"image"
]
] | econ |
According to inequality (5), which denominator appears under (∑√c_i(a_i))^2? | ∑_{i=1}^n α_i | 2406.11528 | [
14
] | [
[
[
281,
1453,
2263,
2912
]
]
] | [
[
[
110.196078,
440.30303,
887.45098,
882.424242
]
]
] | [
[
"image"
]
] | econ |
In the final inequality, E_Fw[y − w(y)] is bounded by what expression that is set equal to u(α*)/−ln(1−α*)? | (1 − α_w) · e*(α_w) | 2406.11528 | [
10
] | [
[
[
290,
965,
2254,
2639
]
]
] | [
[
[
113.72549,
292.424242,
883.921569,
799.69697
]
]
] | [
[
"image"
]
] | econ |
In (P1) the function e(α) appears; in the proof’s argmax definition, what ordered pair defines a(α), and which symbol replaces e(α) as its first component? | a(α)=(F(α), c(α)); F(α). | 2406.11528 | [
7
] | [
[
[
273,
2216,
2257,
3038
]
]
] | [
[
[
107.058824,
671.515152,
885.098039,
920.606061
]
]
] | [
[
"image"
]
] | econ |
In the expression defining V_P(A_0), what is the subscript on y inside the expectation and is A constrained to be a superset or subset of A_0 in the inner minimization? | y_1; superset (A ⊇ A_0). | 2406.11528 | [
18
] | [
[
[
308,
2034,
2248,
2278
]
]
] | [
[
[
120.784314,
616.363636,
881.568627,
690.30303
]
]
] | [
[
"image"
]
] | econ |
In the displayed maximization defining λ*, how many times does the fraction λ/(λ+1) appear in the objective, and where relative to the expectations E_F0[·]? | Twice: before E_F0[y1] and inside b preceding E_F0[y]. | 2406.11528 | [
18
] | [
[
[
296,
2313,
2260,
2592
]
]
] | [
[
[
116.078431,
700.909091,
886.27451,
785.454545
]
]
] | [
[
"image"
]
] | econ |
In Lemma 4.3, which relation holds between A and A0, and what exact expression (including the sign) upper-bounds any contract’s payoff with respect to A? | A ⊇ A0; u(α*)/−ln(1−α*). | 2406.11528 | [
6
] | [
[
[
305,
1802,
2260,
1958
]
]
] | [
[
[
119.607843,
546.060606,
886.27451,
593.333333
]
]
] | [
[
"image"
]
] | econ |
In program (P3), over which exact set of α are both the integral and the universal quantifier taken? | α ∈ R_+^n with ∑_{i=1}^n α_i ≤ 1. | 2406.11528 | [
13
] | [
[
[
302,
2566,
2257,
3047
]
]
] | [
[
[
118.431373,
777.575758,
885.098039,
923.333333
]
]
] | [
[
"image"
]
] | econ |
In the definition of e*(α), which variable is inside the logarithm, and what exact factor multiplies the logarithm in the denominator? | α*; −(1 − α). | 2406.11528 | [
9
] | [
[
[
281,
1923,
2278,
3026
]
]
] | [
[
[
110.196078,
582.727273,
893.333333,
916.969697
]
]
] | [
[
"image"
]
] | econ |
From the shown outputs, what is p_value (kendall_cor_test with alternative='less') minus the Kendall correlation returned by kendall_cor(x, y)? | 0.8423971 | 2408.09618 | [
9
] | [
[
[
287,
1541,
2266,
2965
]
]
] | [
[
[
112.54902,
466.969697,
888.627451,
898.484848
]
]
] | [
[
"image"
]
] | econ |
In the legend titled 'Implementation', which implementation uses orange circles, and what is the rightmost x-axis tick label? | kendallknight; 100,000 | 2408.09618 | [
7
] | [
[
[
287,
2008,
1961,
2883
]
]
] | [
[
[
112.54902,
608.484848,
769.019608,
873.636364
]
]
] | [
[
"image"
]
] | econ |
Compute (base R time at 20,000 − base R time at 10,000) − (kendallknight time at 20,000 − kendallknight time at 10,000). | 4.055 s | 2408.09618 | [
6
] | [
[
[
302,
2771,
2254,
3009
]
]
] | [
[
[
118.431373,
839.69697,
883.921569,
911.818182
]
]
] | [
[
"table"
]
] | econ |
At what number of observations does kendallknight first exceed base R memory by more than 2.5 MB? | 90,000 | 2408.09618 | [
7
] | [
[
[
281,
1039,
2263,
1843
]
]
] | [
[
[
110.196078,
314.848485,
887.45098,
558.484848
]
]
] | [
[
"table"
]
] | econ |
In the two install commands shown, which uses a dot and which uses an underscore in the function name for CRAN vs GitHub? | CRAN: install.packages; GitHub: install_github. | 2408.09618 | [
9
] | [
[
[
278,
992,
2266,
1300
]
]
] | [
[
[
109.019608,
300.606061,
888.627451,
393.939394
]
]
] | [
[
"image"
]
] | econ |
For a dataset with 20,000 observations, approximately how many times more operations does O(n^2) require than O(n log(n)) to obtain the Kendall’s correlation coefficient? | About 2020 times. | 2408.09618 | [
3
] | [
[
[
290,
2571,
2257,
2800
]
]
] | [
[
[
113.72549,
779.090909,
885.098039,
848.484848
]
]
] | [
[
"text"
]
] | econ |
List the four citation years mentioned, in the order they appear. | 1966, 1999, 2005, 2024 | 2408.09618 | [
3
] | [
[
[
287,
2284,
2260,
2521
]
]
] | [
[
[
112.54902,
692.121212,
886.27451,
763.939394
]
]
] | [
[
"text"
]
] | econ |
According to the text, what is Kendall’s R implementation complexity and the example observation count at which it can be slow? | O(n^2); 10,000 observations or more. | 2408.09618 | [
3
] | [
[
[
299,
1823,
2263,
2237
]
]
] | [
[
[
117.254902,
552.424242,
887.45098,
677.878788
]
]
] | [
[
"text"
]
] | econ |
According to the passage, besides precision, what is not sacrificed? | correct handling of corner cases | 2408.09618 | [
3
] | [
[
[
290,
2841,
2251,
2988
]
]
] | [
[
[
113.72549,
860.909091,
882.745098,
905.454545
]
]
] | [
[
"text"
]
] | econ |
Comparing steps 6 and 7, which tie-count variables are subtracted from m in each case for computing t and r(x, y)'s denominator? | Step 6: t_x, t_y; Step 7: m_x, m_y. | 2408.09618 | [
6
] | [
[
[
361,
725,
2254,
1447
]
]
] | [
[
[
141.568627,
219.69697,
883.921569,
438.484848
]
]
] | [
[
"text"
]
] | econ |
Which circled numbers label, in order: trust service providers authorized by regulators, revoking certificates, integrators, and signature creation? | ①, ②, ③, ④ | 2408.14349 | [
4
] | [
[
[
278,
1685,
2128,
2692
]
]
] | [
[
[
109.019608,
510.606061,
834.509804,
815.757576
]
]
] | [
[
"image"
]
] | cs |
Which single number labels both the APIs→Third-party Platforms link and the SFTP systems→Institutional Platforms link? | 8 | 2408.14349 | [
5
] | [
[
[
322,
728,
2157,
2008
]
]
] | [
[
[
126.27451,
220.606061,
845.882353,
608.484848
]
]
] | [
[
"image"
]
] | cs |
Which selected solution has different validity status between Türkiye and the European Union? | Remote Signature | 2408.14349 | [
3
] | [
[
[
293,
496,
2128,
930
]
]
] | [
[
[
114.901961,
150.30303,
834.509804,
281.818182
]
]
] | [
[
"table"
]
] | cs |
Whose systematization inspired their methodology, and to what was it applied? | Simmons and colleagues; Qualified Electronic Signatures (QES). | 2408.14349 | [
2
] | [
[
[
293,
1465,
2134,
1685
]
]
] | [
[
[
114.901961,
443.939394,
836.862745,
510.606061
]
]
] | [
[
"text"
]
] | cs |
Which three media are listed for creating qualified electronic signatures? | QES tokens, QES in the national IDs, mobile signatures. | 2408.14349 | [
11
] | [
[
[
287,
1054,
2119,
1159
]
]
] | [
[
[
112.54902,
319.393939,
830.980392,
351.212121
]
]
] | [
[
"text"
]
] | cs |
According to the paragraph and the caption, how many authorized offices does the cited operator have, which operator is it, and from which platform are the screenshots? | 94; Vodafone; e-Devlet | 2408.14349 | [
10
] | [
[
[
299,
554,
2137,
689
]
]
] | [
[
[
117.254902,
167.878788,
838.039216,
208.787879
]
]
] | [
[
"text"
]
] | cs |
The text expands JNLP in two ways; what single-word difference appears between them? | Launch versus Launching | 2408.14349 | [
14
] | [
[
[
290,
1893,
2131,
2665
]
]
] | [
[
[
113.72549,
573.636364,
835.686275,
807.575758
]
]
] | [
[
"text"
]
] | cs |
Across the crops, what is the third institution listed, and what acronym labels the registered email address accessible after QES authentication? | Postal and telegraph organization; KEP. | 2408.14349 | [
13
] | [
[
[
293,
2046,
2122,
2154
]
]
] | [
[
[
114.901961,
620,
832.156863,
652.727273
]
]
] | [
[
"text"
]
] | cs |
In the related work, what reference numbers are used for Truong and Minh-Tuan in their initial mention and in the later comparison? | 15 and 14 | 2408.14349 | [
6
] | [
[
[
293,
601,
2128,
1908
]
]
] | [
[
[
114.901961,
182.121212,
834.509804,
578.181818
]
]
] | [
[
"text"
]
] | cs |
According to the text, which step applies cognitive walkthroughs, and what specific relationship between ELS and QES is suggested? | Third step; ELS is a subset of QES. | 2408.14349 | [
2
] | [
[
[
287,
1699,
2128,
1846
]
]
] | [
[
[
112.54902,
514.848485,
834.509804,
559.393939
]
]
] | [
[
"text"
]
] | cs |
In the Lagrangian L(α_c, κ_{i,c}, λ), what are the summation ranges for the first double sum and for the λ-parentheses term, respectively? | i=1..n, c=1..C; c=1..C | 2409.08773 | [
40
] | [
[
[
449,
1344,
2105,
2818
]
]
] | [
[
[
176.078431,
407.272727,
825.490196,
853.939394
]
]
] | [
[
"image"
]
] | econ |
In the 'Cluster 1' plot, what are the highest y-axis tick value and the leftmost x-axis tick label shown? | 30 and 0.75 | 2409.08773 | [
28
] | [
[
[
463,
684,
1212,
1444
]
]
] | [
[
[
181.568627,
207.272727,
475.294118,
437.575758
]
]
] | [
[
"image"
]
] | econ |
Between n=400 and n=800 in panel (a), which has the lower minimum Rand Index, and according to panel (b) what number of clusters is selected for that same n? | n=400; 4 | 2409.08773 | [
26
] | [
[
[
458,
957,
2105,
2031
]
]
] | [
[
[
179.607843,
290,
825.490196,
615.454545
]
]
] | [
[
"image"
]
] | econ |
What is the difference between Cluster 4’s highest Treatment tick and Cluster 5’s lowest Treatment tick? | 0.25 | 2409.08773 | [
45
] | [
[
[
458,
1526,
2099,
2401
]
]
] | [
[
[
179.607843,
462.424242,
823.137255,
727.575758
]
]
] | [
[
"image"
]
] | econ |
At approximately Treatment=6 in the plot, which estimator has the highest predicted Outcome among the four curves? | Unweighted Est. | 2409.08773 | [
7
] | [
[
[
443,
927,
2093,
2061
]
]
] | [
[
[
173.72549,
280.909091,
820.784314,
624.545455
]
]
] | [
[
"image"
]
] | econ |
In the CI-DRF minimization problem, what is the exact constraint on κ and which numbered equations must hold for optimal solutions? | ∑_{c=1}^C κ_{i,c}=1; Equations (3) and (4). | 2409.08773 | [
18
] | [
[
[
458,
1553,
2093,
2292
]
]
] | [
[
[
179.607843,
470.606061,
820.784314,
694.545455
]
]
] | [
[
"image"
]
] | econ |
In Figure B.16, which estimator’s partitions are evaluated by the Rand Index, and what single number of clusters is selected for both n=400 and n=800? | CI-DRF; 3 | 2409.08773 | [
46
] | [
[
[
466,
1294,
2093,
2354
]
]
] | [
[
[
182.745098,
392.121212,
820.784314,
713.333333
]
]
] | [
[
"image"
]
] | econ |
Comparing the OLS summation and matrix forms, which index appears on κ in κ_{i,c}^{(init)} but not in (Z' κ_c^{(init)} Z)^{-1} Z κ_c^{(init)} y? | i | 2409.08773 | [
19
] | [
[
[
449,
1808,
2084,
2560
]
]
] | [
[
[
176.078431,
547.878788,
817.254902,
775.757576
]
]
] | [
[
"image"
]
] | econ |
At EU funds per capita ≈400, order the three ADRF estimates from highest to lowest growth rate. | DCOW-weighted (green) > HI2004 (red) > Unweighted (purple) | 2409.08773 | [
31
] | [
[
[
437,
1658,
2087,
2739
]
]
] | [
[
[
171.372549,
502.424242,
818.431373,
830
]
]
] | [
[
"image"
]
] | econ |
Using the caption and the plot, which function is described as average, and what symbol labels it on the curve? | Dose-response; µ(t) | 2409.08773 | [
13
] | [
[
[
551,
1186,
1946,
1984
]
]
] | [
[
[
216.078431,
359.393939,
763.137255,
601.212121
]
]
] | [
[
"image"
]
] | econ |
In Assumption 5, En[Z^T M_X ? Y] ≠ 0 appears in both parts. What symbol replaces '?' in parts i and ii, respectively? | G in i, A in ii. | 2410.14317 | [
17
] | [
[
[
287,
1080,
1852,
1591
]
]
] | [
[
[
112.54902,
327.272727,
726.27451,
482.121212
]
]
] | [
[
"image"
]
] | econ |
According to Assumption 1 (Exchangeability), on which set is the permutation π defined? | {1,…,n} ∖ {i} | 2410.14317 | [
6
] | [
[
[
293,
2193,
1652,
2574
]
]
] | [
[
[
114.901961,
664.545455,
647.843137,
780
]
]
] | [
[
"image"
]
] | econ |
In the restricted model, what are β_{k,d} (for k<d), β_{d,d}, and the d range these restrictions apply to? | β_-max/(d-1); β_{d,d}=β_max; d=1,...,5. | 2410.14317 | [
21
] | [
[
[
296,
2480,
2269,
2953
]
]
] | [
[
[
116.078431,
751.515152,
889.803922,
894.848485
]
]
] | [
[
"image"
]
] | econ |
Which equation number labels the full-rank matrix, and to which corollary are the authors’ different conditions related and in which assumption are they given? | Equation (9), Corollary 1, Assumption 4. | 2410.14317 | [
13
] | [
[
[
284,
1570,
2260,
2563
]
]
] | [
[
[
111.372549,
475.757576,
886.27451,
776.666667
]
]
] | [
[
"image"
]
] | econ |
According to Theorem 2, which assumptions are invoked, and in the TSLS estimator expression, which summation uniquely has a 1/√n coefficient instead of 1/n? | Assumptions 2-4 and 6; the sum ∑ z_i Y_i. | 2410.14317 | [
26
] | [
[
[
287,
1873,
1858,
2909
]
]
] | [
[
[
112.54902,
567.575758,
728.627451,
881.515152
]
]
] | [
[
"image"
]
] | econ |
In the theorem shown, which theorem establishes identification, what lower bound on n is given, and what is the equation number of the moment condition? | Theorem 1; n ≥ d(d+1)/2 + l; (11). | 2410.14317 | [
15
] | [
[
[
293,
2345,
2284,
2888
]
]
] | [
[
[
114.901961,
710.606061,
895.686275,
875.151515
]
]
] | [
[
"image"
]
] | econ |
Which identity-matrix symbol differs between the inverses defining θ and θ̃ rows? | I for θ; I_n for θ̃. | 2410.14317 | [
32
] | [
[
[
290,
1520,
2257,
2624
]
]
] | [
[
[
113.72549,
460.606061,
885.098039,
795.151515
]
]
] | [
[
"image"
]
] | econ |
In the definition of β_ij(π), when A_ij = 1, by what exact summation equality is k determined? | ∑_{j'=1}^{π^{-1}(j)} A_{iπ(j')} = k | 2410.14317 | [
31
] | [
[
[
305,
1699,
2287,
2900
]
]
] | [
[
[
119.607843,
514.848485,
896.862745,
878.787879
]
]
] | [
[
"image"
]
] | econ |
According to the text under Eq. (2), from which set and by what ordering is Ỹi,k selected? | k-th lowest of {Y_j: j is a peer of i} | 2410.14317 | [
3
] | [
[
[
302,
939,
2257,
1732
]
]
] | [
[
[
118.431373,
284.545455,
885.098039,
524.848485
]
]
] | [
[
"image"
]
] | econ |
After aggregating across k and i, which exact constant multiplies ||y − y'||∞ in the final bound on ||g(y) − g(y')||∞? | β̄ | 2410.14317 | [
30
] | [
[
[
287,
733,
2260,
2680
]
]
] | [
[
[
112.54902,
222.121212,
886.27451,
812.121212
]
]
] | [
[
"image"
]
] | econ |
Which variable has negative dy/dx in both services and shifts from 1% significance for shelling to 10% for ploughing? | Farm size; negative in both. | 2411.15797 | [
11
] | [
[
[
299,
1001,
2269,
2375
]
]
] | [
[
[
117.254902,
303.333333,
889.803922,
719.69697
]
]
] | [
[
"table"
]
] | econ |
Which constraint ranks last among 'No' respondents but second overall? | Machines not available | 2411.15797 | [
14
] | [
[
[
302,
1738,
2222,
2783
]
]
] | [
[
[
118.431373,
526.666667,
871.372549,
843.333333
]
]
] | [
[
"table"
]
] | econ |
Comparing ploughing and shelling, what are the absolute differences in their 'Yes' and 'No' proportions, respectively? | 0.01 and 0.01 | 2411.15797 | [
10
] | [
[
[
290,
484,
2301,
1652
]
]
] | [
[
[
113.72549,
146.666667,
902.352941,
500.606061
]
]
] | [
[
"image"
]
] | econ |
Compute |(mean No − mean Yes) − reported Diff| from the table. | 0.0001 | 2411.15797 | [
13
] | [
[
[
252,
1450,
2272,
1967
]
]
] | [
[
[
98.823529,
439.393939,
890.980392,
596.060606
]
]
] | [
[
"table"
]
] | econ |
What is the total number of respondents from the communities with hyphenated names? | 31 | 2411.15797 | [
5
] | [
[
[
293,
2460,
2269,
2953
]
]
] | [
[
[
114.901961,
745.454545,
889.803922,
894.848485
]
]
] | [
[
"table"
]
] | econ |
In the table, which two independent variables have their expected-sign symbols placed before the variable names rather than in the Expected Sign column? | Farming_experience and Profit_per_acre | 2411.15797 | [
7
] | [
[
[
293,
962,
2334,
2143
]
]
] | [
[
[
114.901961,
291.515152,
915.294118,
649.393939
]
]
] | [
[
"table"
]
] | econ |
According to the paragraph, which metric reveals the disparity between operational costs and revenue, and which financial indicator's mean is compared using a t-test? | Gross Profit Margin; net profit. | 2411.15797 | [
7
] | [
[
[
270,
2207,
2251,
2451
]
]
] | [
[
[
105.882353,
668.787879,
882.745098,
742.727273
]
]
] | [
[
"text"
]
] | econ |
By how many points does the mean score of environmental regulations/concerns exceed that of financial issues? | 3.75 | 2411.15797 | [
14
] | [
[
[
293,
936,
2266,
1271
]
]
] | [
[
[
114.901961,
283.636364,
888.627451,
385.151515
]
]
] | [
[
"text"
]
] | econ |
According to the excerpt, who is credited with propounding the Profit Maximization Theory and in what year? | Adam Smith, 2000 | 2411.15797 | [
3
] | [
[
[
299,
2442,
2257,
2856
]
]
] | [
[
[
117.254902,
740,
885.098039,
865.454545
]
]
] | [
[
"text"
]
] | econ |
According to the survey, how many of how many respondents used tractor services, and what percentage was this? | 322 of 359 respondents, 89.7%. | 2411.15797 | [
11
] | [
[
[
293,
2410,
2266,
2695
]
]
] | [
[
[
114.901961,
730.30303,
888.627451,
816.666667
]
]
] | [
[
"text"
]
] | econ |
According to Corollary 2, which two conditions are satisfied besides being a homogeneous, neutral, pairwise Condorcet extension? | optimist participation and positive responsiveness | 2411.19857 | [
17
] | [
[
[
277,
1616,
2200,
1732
]
]
] | [
[
[
111.648529,
460.661345,
886.739218,
493.72862
]
]
] | [
[
"image"
]
] | econ |
Which theorem number, and under which participation condition, states that a homogeneous Condorcet extension is a refinement of maximin? | Optimist participation; Theorem 5. | 2411.19857 | [
14
] | [
[
[
287,
1164,
2197,
1282
]
]
] | [
[
[
115.679162,
331.812999,
885.530028,
365.450399
]
]
] | [
[
"image"
]
] | econ |
According to the proof, who is the Condorcet winner in P1, and what is f(P1+P2) after rearranging voters using anonymity? | c; {a} | 2411.19857 | [
10
] | [
[
[
287,
1572,
2197,
3195
]
]
] | [
[
[
115.679162,
448.118586,
885.530028,
910.775371
]
]
] | [
[
"image"
]
] | econ |
Across all blocks (Case 1, Case 2, and the unlabeled ones), what is the difference between the maximum number beside b and the smallest nonzero number beside c? | 11 | 2411.19857 | [
12
] | [
[
[
596,
1223,
1956,
1482
]
]
] | [
[
[
240.225715,
348.631699,
788.391778,
422.462942
]
]
] | [
[
"image"
]
] | econ |
According to the excerpt, which lemma specifies when f equals maximin, and what two conditions on f are required? | Lemma 4; continuous and a refinement of maximin. | 2411.19857 | [
18
] | [
[
[
290,
2833,
2200,
2943
]
]
] | [
[
[
116.888351,
807.582668,
886.739218,
838.939567
]
]
] | [
[
"image"
]
] | econ |
In the definition of f_defensible(P), which variable is universally quantified and which is existentially quantified? | y universal; z existential. | 2411.19857 | [
13
] | [
[
[
293,
2050,
2190,
2549
]
]
] | [
[
[
118.097541,
584.378563,
882.708585,
726.624857
]
]
] | [
[
"image"
]
] | econ |
Across the first rows of profiles P7 and P8 (from P1+P2 and P1+P3), how many colored c's appear in total? | 5 | 2411.19857 | [
9
] | [
[
[
283,
586,
2215,
2980
]
]
] | [
[
[
114.066909,
167.04675,
892.785167,
849.486887
]
]
] | [
[
"image"
]
] | econ |
In Corollary 3, besides optimist participation, which additional property ensures Nanson’s rule uniquely among homogeneous, neutral, pairwise strong Condorcet extensions? | tie-break positive responsiveness | 2411.19857 | [
18
] | [
[
[
284,
2259,
2194,
2359
]
]
] | [
[
[
114.469972,
643.95667,
884.320838,
672.462942
]
]
] | [
[
"image"
]
] | econ |
In Theorem 6, which specific theorem and proposition from Holliday and Pacuit (2023c) are cited? | Theorem 3.8 and Proposition 3.10 | 2411.19857 | [
16
] | [
[
[
284,
736,
2190,
901
]
]
] | [
[
[
114.469972,
209.806157,
882.708585,
256.841505
]
]
] | [
[
"image"
]
] | econ |
Which lemma number states that Leximin is the only neutral and pairwise refinement of maximin that satisfies positive responsiveness? | Lemma 2. | 2411.19857 | [
16
] | [
[
[
287,
2147,
2190,
2247
]
]
] | [
[
[
115.679162,
612.029647,
882.708585,
640.535918
]
]
] | [
[
"image"
]
] | econ |
End of preview. Expand in Data Studio
SciEGQA Bench
SciEGQA is a scientific document question answering and reasoning dataset with semantic evidence grounding, where supporting evidence is represented as semantically coherent document regions annotated with bounding boxes.
Quick start
import json
from pathlib import Path
from PIL import Image, ImageDraw
with open("SciEGQA_Bench.jsonl", "r", encoding="utf-8") as f:
sample = json.loads(next(f))
for page, boxes in zip(sample["evidence_page"], sample["bbox"]):
page_path = (
Path(sample["category"])
/ sample["doc_name"]
/ f"{sample['doc_name']}_{page}.png"
)
image = Image.open(page_path).convert("RGB")
draw = ImageDraw.Draw(image)
for xmin, ymin, xmax, ymax in boxes:
draw.rectangle((xmin, ymin, xmax, ymax), outline="red", width=5)
image.show()
Notes
evidence_pageis 1-based and directly matches the page number in the PNG file name.- Samples may reference either one page or two pages.
- Bounding boxes are provided in both absolute pixel coordinates (
bbox) and normalized coordinates (rel_bbox). - The benchmark is intended for evaluating both answer quality and evidence grounding quality.
- Downloads last month
- 163