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

SAT Solving on 100 SAT Benchmarks (50 SAT, 50 UNSAT) (test)

52Pass@1

unit

Updated 15d ago

Evaluation Results

Method	Links
unit 2026.05		52	53	53	53	53	26.4	25.2	51.6	0.5	1.2	0.033	-
march_cu 2026.05		51	52	52	52	52	26.6	25	51.6	0.5	1.6	0.003	-
heule_schur 2026.05		50	52	52	52	52	26.4	25	51.4	0.9	1.4	0.025	-
Qwen3-4B-SFT 2026.05		49	49	50	51	51	22.6	23.4	46	2.3	0.8	0.242	-
Claude-Sonnet-4 2026.05		47	50	50	50	50	23.8	23	46.8	1.5	0.8	0.294	-
Qwen3-32B 2026.05		46	48	49	49	49	21	22	43	2.4	1	0.089	-
Qwen3-4B-SFT-DPO 2026.05		46	51	51	53	53	23.8	23.6	47.4	0.9	0.2	0.778	-
Claude-3.7-Sonnet 2026.05		44	48	48	48	48	22	22.6	44.6	0.5	0.6	0.426	-
GPT-OSS-120B 2026.05		43	45	45	46	47	20.8	22	42.8	0.8	1.2	0.033	-
Qwen3-4B-DPO 2026.05		42	45	45	46	47	19.8	21.8	41.6	1.1	2	0.003	-
Qwen3-4B 2026.05		40	46	46	46	46	21.2	21.6	42.8	1.9	0.4	0.689	-