List your company
ResearchFormerly WizWand

Preference Prediction on PPE Preference (test)

79.8Preference ScoreSkywork-Reward-V2-Llama-3.1-8B-40M
58.37663.93869.575.062Jul 2, 2025
Updated 8mo ago

Evaluation results

MethodLinks
Skywork-Reward-V2-Llama-3.1-8B-40M2025.0779.8
Skywork-Reward-V2-Llama-3.1-8B2025.0777.3
Skywork-Reward-V2-Qwen3-8B2025.0770.6
Skywork-Reward-V2-Qwen3-4B2025.0769.5
Skywork-Reward-V2-Llama-3.2-3B2025.0769.1
GPT-4oModel Type=LLM-as-a-Judge2025.0767.7
Skywork-Reward-V2-Qwen3-1.7B2025.0767.6
Claude-3.5-SonnetModel Type=LLM-as-a-Judge2025.0767.3
DeepSeek-GRM-27B (w/ MetaRM)Model Type=Generative Reward Model2025.0767.2
J1-Llama-70B (Maj@32)Model Type=LLM-as-a-Judge, Sampling Strategy=Maj@322025.0767
Skywork-Reward-V2-Llama-3.2-1B2025.0766.6
J1-Llama-70BModel Type=LLM-as-a-Judge2025.0766.3
DeepSeek-GRM-27BModel Type=Generative Reward Model2025.0765.3
Skywork-Reward-V2-Qwen3-0.6B2025.0765.3
Llama-3.1-Nemotron-70BModel Type=Open Reward Model2025.0764.2
INF-ORM-Llama3.1-70BModel Type=Open Reward Model2025.0764.2
Skywork-Reward-Gemma-2-27B-v0.2Model Type=Open Reward Model2025.0763.6
LDL-Reward-Gemma-2-27B-v0.1Model Type=Open Reward Model2025.0762.4
Skywork-Reward-Llama-3.1-8B-v0.2Model Type=Open Reward Model2025.0762.2
Internlm2-20b-rewardModel Type=Open Reward Model2025.0761
ArmoRM-Llama3-8B-v0.1Model Type=Open Reward Model2025.0760.6
J1-Llama-8B (Maj@32)Model Type=LLM-as-a-Judge, Sampling Strategy=Maj@322025.0760.6
J1-Llama-8BModel Type=LLM-as-a-Judge2025.0760.3
Llama-3-OffsetBias-RM-8BModel Type=Open Reward Model2025.0759.2