Share your thoughts, 1 month free Claude Pro on usSee more

Table Reasoning on WTQ + FeTaQA + TabFact Combined (test)

64.7F1 Score

RSAT

Updated 1mo ago

Evaluation Results

Method	Links
RSAT 2026.04		64.7	100	97.2	100	100
RSAT 2026.04		61.9	99.2	97.7	99.2	99.2
RSAT 2026.04		59.2	99.9	94.6	99.6	100
RSAT 2026.04		59.2	99.3	73.5	86.5	99.4
SFT 2026.04		57.6	100	23.4	88.8	100
SFT 2026.04		55.5	99.6	28.8	83	99.8
SFT 2026.04		54.6	98.4	26.9	82.2	98.4
RSAT 2026.04		53.7	97.2	48	96.7	97.2
SFT 2026.04		53.1	99.6	21.3	84.8	99.8
RSAT 2026.04		52.4	99.6	84.7	99	99.8
SFT 2026.04		40.3	94.7	19.2	78.7	95.2
SFT 2026.04		37.1	99.5	14.9	91.8	99.8
Zero-Shot 2026.04		27.7	56.5	8.6	50.1	56.6
Zero-Shot 2026.04		20.9	39.1	4	38.9	43.6
Zero-Shot 2026.04		18.1	29.6	4	36.4	39.8
Post-Hoc 2026.04		11.5	18.2	3.2	20.9	24.4
Zero-Shot 2026.04		10	32.6	0.8	28	33.4
Zero-Shot 2026.04		9.9	36.8	1.5	33.6	36.8
Post-Hoc 2026.04		8.8	14	2.7	11.7	14
Post-Hoc 2026.04		6.6	6.4	0.4	18.4	19.8
Zero-Shot 2026.04		5.4	16.4	0.6	14.2	16.4
Post-Hoc 2026.04		3.5	5.6	0.5	8.5	9
Post-Hoc 2026.04		1.3	4	0.6	3.8	4
Post-Hoc 2026.04		0.2	0	0	0.4	0.4