paper-with-me

홈 › Papers

Reliable to Expressive: A Curriculum for Rubric-Following Safety Judges

2026-06-08 · Yongtaek Lim, Hyeji Choi, Minwoo Kim arxiv

Safety judges are increasingly deployed to evaluate model outputs against evolving criteria, yet recent meta-evaluation work shows they remain brittle under prompt and rubric variation, with false negative-rate swings of up to 0.24 reported for stylistic perturbations alone. We argue that safety judgment is fundamentally a rubric-following problem: a robust judge must apply the given evaluation criteria consistently across rubric formulations rather than memorize one specific template. We propose a training strategy that combines (i) instance-conditioned dynamic rubrics generated from prompt-response-label triples to expose the judge to the variability of evaluation criteria, and (ii) a reliable-to-expressive curriculum that begins with clean fixed-rubric supervision and progressively introduces noisier dynamic-rubric data. We evaluate on a single human-labeled set under three contrasting rubric prompts (HarmBench-style, ShieldGemma-style, and a domain-specific rubric). Our 12B curriculum judge achieves 94.12-94.88% accuracy across the three rubrics with a cross-rubric range of only 0.76, outperforming general-purpose LLMs, dedicated safety classifiers, and reasoning-oriented judges up to 30B in both peak accuracy and stability. An ablation shows that naively mixing dynamic rubrics into SFT increases cross rubric variance (1.44 -> 3.60); only the curriculum schedule recovers and improves on the fixed rubric baseline (variance 0.76).

📄 PDF Abstract BibTeX arXiv:2606.09165

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

AdvancedIF: Rubric-Based Benchmarking and Reinforcement Learning for Advancing LLM Instruction Following

2025-11-13 · Yun He, Wenzhe Li, Hejia Zhang, Songlin Li 외 arxiv

Recent progress in large language models (LLMs) has led to impressive performance on a range of tasks, yet advanced instruction following (IF)-especially for complex, multi-turn, and system-prompted instructions-remains …

Reinforcement LearningInstruction Following

RubricEval: A Rubric-Level Meta-Evaluation Benchmark for LLM Judges in Instruction Following

2026-03-26 · Tianjun Pan, Xuan Lin, Wenyan Yang, Qianyu He 외 arxiv

Rubric-based evaluation has become a prevailing paradigm for evaluating instruction following in large language models (LLMs). Despite its widespread use, the reliability of these rubric-level evaluations remains unclear…

Instruction Following

CITING: Large Language Models Create Curriculum for Instruction Tuning

2023-10-04 · Tao Feng, Zifeng Wang, Jimeng Sun

The recent advancement of large language models (LLMs) has been achieved through a combo of instruction tuning and human alignment. However, building manually crafted instruction datasets and performing human alignment b…

AMARIS: A Memory-Augmented Rubric Improvement System for Rubric-Based Reinforcement Learning

2026-05-18 · Peilin Wu, Xinlu Zhang, Kun Wan, Wentian Zhao 외 arxiv

Rubric-based reward shaping provides interpretable and editable reward signals for fine-tuning LLMs via reinforcement learning (RL), but existing adaptive rubric methods typically update criteria from local evidence such…

Reinforcement LearningInstruction Following

RuCL: Stratified Rubric-Based Curriculum Learning for Multimodal Large Language Model Reasoning

2026-02-25 · Yukun Chen, Jiaming Li, Longze Chen, Ze Gong 외 arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a prevailing paradigm for enhancing reasoning in Multimodal Large Language Models (MLLMs). However, relying solely on outcome supervision risks reward …

Reinforcement LearningLogical ReasoningVisual Reasoning