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

Math Reasoning on GSM8K (pass@4)

95.83Pass@4 Accuracy

JURY-RL

Updated 1mo ago

Evaluation Results

Method	Links
JURY-RL 2026.04		95.83	64.04
LLM-KD 2026.04		95.75	63.54
CoReward 2026.04		95.6	59.99
LLM-as-a-Judge 2026.04		95.38	53.99
GT-Reward 2026.04		95.22	62.48
Majority-Voting 2026.04		95	61.52
Self-Certainty 2026.04		93.86	60.33
Before RL 2026.04		93.78	56.77
Entropy 2026.04		93.78	62.62
LLM-KD 2026.04		93.25	53.51
GT-Reward 2026.04		92.42	55.36
JURY-RL 2026.04		92.42	59.41
Majority-Voting 2026.04		92.34	52.51
CoReward 2026.04		92.04	53.4
LLM-as-a-Judge 2026.04		91.74	50.35
Majority-Voting 2026.04		90.98	42.2
LLM-KD 2026.04		90.96	-
Self-Certainty 2026.04		90.83	55.17
GT-Reward 2026.04		90.83	-
Entropy 2026.04		90.75	54.86
Majority-Voting 2026.04		90.52	-
CoReward 2026.04		90.45	43.51
JURY-RL 2026.04		90.35	-
CoReward 2026.04		90.16	-
GT-Reward 2026.04		90.14	45.46
JURY-RL 2026.04		90.07	48.48
LLM-KD 2026.04		89.99	46.21
LLM-as-a-Judge 2026.04		89.84	45.93
LLM-as-a-Judge 2026.04		89.73	-
Before RL 2026.04		89.31	49.59
Self-Certainty 2026.04		88.21	-
Entropy 2026.04		87.85	-
Before RL 2026.04		86.88	46.26
Self-Certainty 2026.04		84.69	42.12
Entropy 2026.04		84.46	42.77
JURY-RL 2026.04		82.58	-
GT-Reward 2026.04		82.01	-
LLM-KD 2026.04		81.97	-
LLM-as-a-Judge 2026.04		81.84	-
Majority-Voting 2026.04		81.08	-
CoReward 2026.04		81.07	-
CoReward 2026.04		80	-
LLM-KD 2026.04		79.72	-
Entropy 2026.04		79.45	-
JURY-RL 2026.04		78.96	-
Majority-Voting 2026.04		78.64	-
GT-Reward 2026.04		78.6	-
LLM-as-a-Judge 2026.04		78.51	-
Self-Certainty 2026.04		73.71	-
Self-Certainty 2026.04		73.52	-
Before RL 2026.04		71	-
Before RL 2026.04		68.78	-
Entropy 2026.04		68.57	-
Before RL 2026.04		64.2	-