评判评判者：Rubrics as Rewards 中的置信度与分歧

View PDF HTML (experimental)

Abstract:Rubrics as Rewards (RaR) extends reinforcement learning to non-verifiable domains by using an LLM judge to score responses against a structured rubric. Prior work has focused on improving the rubric or strengthening the single judge, with comparatively little attention to how different single or multiple judges might be used during RaR training. We first show that judge identity alone produces a substantial spread in scores when training Qwen3-4B-Base on RaR-Medicine and RaR-Science across five frontier and open-weight judges. Building on this, we propose RaRE (Rubrics as Rewards with Ensembles), a novel judge ensemble strategy for RaR with two mechanisms: a soft-Likert reward that uses judge logprobs to weight more confident judges more heavily, and a drop-on-disagreement mask that excludes high-disagreement rollouts from the GRPO update. We find that this design is critical: while a naive average of judge scores fails to outperform the best single judge, RaRE outperforms every single-judge RaR variant on both HealthBench and GPQA Diamond, with per-theme analysis showing that it inherits the complementary strengths of its constituent judges. Our findings establish judge ensembling as a compelling axis for improving RaR.

Comments:	Project page: this https URL Code: this https URL
Subjects:	Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
Cite as:	arXiv:2610.21179 [cs.LG]
	(or arXiv:2610.21179v1 [cs.LG] for this version)
	https://doi.org/10.48550/arXiv.2610.21179 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Anne Harrington [view email]
[v1] Thu, 15 Oct 2026 03:35:00 UTC (1,346 KB)