Safety
From Preferences to Principles: Rubric-Based Alignment for Grounded Knowledge Answers
Designing effective reward signals for open-domain question answering is challenging because high-quality responses must simultaneously satisfy multiple aspects of answer quality that are difficult to
Designing effective reward signals for open-domain question answering is challenging because high-quality responses must simultaneously satisfy multiple aspects of answer quality that are difficult to capture with a holistic scalar objective. We introduce a rubric-based reward framework that generates query-specific rubrics grounded in retrieved evidence and decomposed into multiple quality dimensions, providing fine-grained supervision during post-training. Averaged across three evaluation axes (composition, grounding, and instruction-following), our approach improves over the…
Related
- EvoRubrics: Dynamic Rubrics as Rewards via Adversarial Co-Evolution for LLM Reinforcement Learning
- AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering
- ConRub-Med: Reinforcement Learning with Consensus Rubrics for Open-Ended Medical Question Answering
Source: Apple ML Research | 2026-08-27