RELAY: Selective Self-Distillation with Adaptive Hand-off for Rubric-Based Reinforcement Learning

arXiv:2610.17623 · cs.LG, cs.AI, cs.CL · Submitted 2026-10-13 · Read on arXiv

cs.LG, cs.AI, cs.CL

Submitted: 2026-10-13

Updated: 2026-10-14

Comments: 17 pages, 8 figures, 13 tables

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Distillation — 1× in this paper · explained in 2 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Self-Distillation — 1× in this paper · explained in 7 episodes

Sources

• HealthBench: Evaluating Large Language Models Towards Improved Human Health
• Rubrics as Privileged Information for Open-Ended Generation
• Rethinking Reward Supervision: Rubric-Conditioned Self-Distillation
• Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains
• Distill Where You Fail: Recovering Learning Signals of Negative RL-Groups from Adaptive Teacher Guidance
• Reinforcement Learning via Self-Distillation
• Rebellious Student: Reversing Teacher Signals for Reasoning Exploration with Self-Distilled RLVR
• Alternating Reinforcement Learning with Contextual Rubric Rewards: Beyond the Scalarization Strategy
• Sequential Beats Joint: On the Interplay between On-Policy Distillation and RLVR
• Unifying Group-Relative and Self-Distillation Policy Optimization via Sample Routing
• RubricHub: A Comprehensive and Highly Discriminative Rubric Dataset via Automated Coarse-to-Fine Generation
• Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
• Self-Distilled Policy Gradient
• Reward Hacking in Rubric-Based Reinforcement Learning
• RLCSD: Reinforcement Learning with Contrastive On-Policy Self-Distillation
• Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers
• Online Rubrics Elicitation from Pairwise Comparisons
• Proximal Policy Optimization Algorithms
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• A Survey of On-Policy Distillation for Large Language Models

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks