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

Reward Modeling on JudgeBench

93.3Accuracy

OpenRS

Updated 1d ago

Evaluation Results

Method
OpenRS 2026.02	93.3	86.8
OpenRS 2026.02	91.6	89.4
OpenRS 2026.02	89.4	86.7
OpenRS 2026.02	89.1	86.2
OpenRS 2026.02	86.4	84.9
Skywork-Reward-V2-Llama-3.1-8B-40M 2025.07	83.4	-
RationaleRM 2026.02	82	-
Skywork-Reward-V2-Llama-3.1-8B 2026.02	80	84.3
Skywork-Reward-V2-Llama-3.1-8B 2025.07	80	-
Skywork-Reward-V2 2025.07	77.9	-
Skywork-Reward-V2 2025.07	76	-
Auto-Rubric 2026.02	74.3	-
Skywork-Reward-V2 2025.07	73.5	-
Skywork-Reward-V2-Qwen3-8B 2025.07	73.4	-
Skywork-Reward-V2-Qwen3-1.7B 2025.07	72.9	-
CAMEL-Reflection 2026.02	71.6	-
SAVE 2026.05	71.4	-
Skywork-Reward-V2 2025.07	71.1	-
INF-ORM-Llama3.1-70B 2026.02	70.2	-
INF-ORM-Llama3.1-70B 2026.02	70.2	71.7
INF-ORM-Llama3.1-70B 2026.02	70.2	-
INF-ORM-Llama3.1-70B 2025.07	70.2	-
INF-ORM-Llama3.1-70B 2025.07	70.2	-
Athene-RM-8B 2026.02	70.1	-
Skywork-Reward-V2 2025.07	70	-
Continual Offline Training RM 2026.05	70	-
SAVE 2026.05	70	-
Skywork-Reward-V2-Qwen3-4B 2025.07	69.5	-
Ray2333/GRM-gemma2-2B-rewardmodel-ft 2025.07	69.4	-
Skywork-Reward-V2-Llama-3.2-3B 2025.07	69.2	-
Skywork-Reward-V2-Llama-3.2-3B 2026.05	69.2	-
CAMEL 2026.02	69.1	-
SAVE (w/o Curriculum Mechanism) 2026.05	69.1	-
SAVE (w/o Curriculum Mechanism) 2026.05	68.9	-
SAVE (w/o Policy Model Optimization) 2026.05	68.3	-
SAVE (w/o Policy Model Optimization) 2026.05	68.3	-
HL-BT 2026.05	67.7	-
HL-BT 2026.05	67.7	-
Skywork-Reward-V2-Qwen3-0.6B 2025.07	67.6	-
Skywork-Reward-Gemma-2-27B-v0.2 2026.02	66.5	-
Skywork-Reward-Gemma-2-27B-v0.2 2026.02	66.5	67.3
Skywork-Reward-Gemma-2-27B-v0.2 2026.02	66.5	-
Skywork-Reward-Gemma-2-27B-v0.2 2025.07	66.5	-
WILDREWARD-8B 2026.02	66	-
Llama-3.1-Nemotron-70B 2026.02	65.8	-
Llama-3.1-Nemotron-70B 2026.02	65.8	70.5
Llama-3.1-Nemotron-70B 2026.02	65.8	-
Llama-3.1-Nemotron-70B 2025.07	65.8	-
CAMEL-Fast 2026.02	65.2	-
Skywork-Reward-Gemma-2-27B 2025.07	65.2	-
Skywork-Reward-V2-Llama-3.2-1B 2025.07	65	-
Claude-3.5-Sonnet 2026.02	64.8	-
RM-R1-Qwen-Instruct-32B 2026.02	64.8	-
Claude-3.5-Sonnet 2025.07	64.8	-
CIR-AMS/BTRM_Qwen2_7b_0613 2025.07	64.5	-
Internlm2-20b-reward 2026.02	64.3	-
Internlm2-20b-reward 2026.02	64.3	64.8
Internlm2-20b-reward 2026.02	64.3	-
Internlm2-20b-reward 2025.07	64.3	-
LDL-Reward-Gemma-2-27B-v0.1 2026.02	64.2	67.4
LDL-Reward-Gemma-2-27B-v0.1 2026.02	64.2	-
LDL-Reward-Gemma-2-27B-v0.1 2025.07	64.2	-
URM-LLaMa-3.1-8B 2025.07	64.1	-
BTRM Qwen2 7b 0613 2025.07	63.8	-
QRM-Llama3.1-8B 2025.07	63.8	-
Starling-RM-34B 2025.07	63.8	-
Llama-3-OffsetBias-RM-8B 2026.02	63.5	-
Llama-3-OffsetBias-RM-8B 2026.02	63.5	65
Llama-3-OffsetBias-RM-8B 2026.02	63.5	-
GRM-gemma2-2B-rewardmodel-ft 2025.07	63.5	-
Llama-3-OffsetBias-RM-8B 2025.07	63.5	-
GRM-llama3-8B-distill 2025.07	63.3	-
Skywork-Reward-Llama-3.1-8B-v0.2 2026.02	62.9	-
Skywork-Reward-Llama-3.1-8B-v0.2 2026.02	62.9	67.1
Skywork-Reward-Llama-3.1-8B-v0.2 2026.02	62.9	-
Skywork-Reward-Llama-3.1-8B-v0.2 2025.07	62.9	-
QRM-Llama3.1-8B-v2 2025.07	62.6	-
GRM-Llama3-8B-rewardmodel-ft 2025.07	62.3	-
RM-Mistral-7B 2025.07	62.1	-
Skywork-Reward-Llama-3.1-8B 2025.07	62	-
RM-R1-Qwen-Instruct-14B 2026.02	61.3	-
WILDREWARD-4B 2026.02	61.1	-
Mean Reward 2026.05	60.9	-
Mean Reward 2026.05	60.9	-
J1-Llama-70B 2026.02	60	-
J1-Llama-70B 2026.02	60	-
J1-Llama-70B 2025.07	60	-
FsfairX-LLaMA3-RM-v0.1 2025.07	59.9	-
GPT-4o 2026.02	59.8	-
GPT-4o 2025.07	59.8	-
ArmoRM-Llama3-8B-v0.1 2026.02	59.7	-
ArmoRM-Llama3-8B-v0.1 2026.02	59.7	63.4
ArmoRM-Llama3-8B-v0.1 2026.02	59.7	-
ArmoRM-Llama3-8B-v0.1 2025.07	59.7	-
Internlm2-7b-reward 2025.07	59.4	-
internlm2-1.8b-reward 2025.07	59	-
Eurus-RM-7b 2025.07	58.4	-
QRM-Gemma-2-27B 2025.07	57.5	-
ConsistRM (Qwen3-8B) 2026.04	56.9	-
Qwen3-8B + TTRL 2026.04	56.8	-

Showing 100 of 117 rows