RM-R1-DeepSeek-Distilled-Qwen-32B

1 benchmarks · 1 observations · 1 sources

Scores at a glance

1 benchmarks · large dot: RM-R1-DeepSeek-Distilled-Qwen-32B; small dots: closest models where reported

RM-Bench ↗83.9%

RM-R1-DeepSeek-Distilled-Qwen-32B · RM-Bench · Overall accuracy · official leaderboard ↗

Closest models by score

Similar positions on shared benchmark leaderboards. Smaller gaps are closer.

Llama-3.3-Nemotron-Super-49B-GenRM + voting@32 →NVIDIA

2.0 percentile-point gap

1 of 1 shared benchmark comparisons

J1-Llama-70B →Meta

3.0 percentile-point gap

1 of 1 shared benchmark comparisons

Llama-3.3-Nemotron-Super-49B-GenRM →NVIDIA

3.0 percentile-point gap

1 of 1 shared benchmark comparisons

How similarity is measured

Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.

Filter comparisons to this model →

Sources

Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.

RM-Bench: Benchmarking Reward Models of Language Models with Subtlety and Style — published result table ↗
Liu, Yantao, Yao, Zijun, Min, Rui, Cao, Yixin, Hou, Lei, Li, Juanzi · 2024

Names used in source tables

RM-R1-DeepSeek-Distilled-Qwen-32B