Shows that under differentiable rollouts with additive noise, actor updates in critic-free RL for LLMs are value-gradient-like in expectation, motivating a decomposition into value signal and reward headroom for when RL is most effective.
arXiv preprint arXiv:2603.01162 , year=
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
citation-role summary
background 4
citation-polarity summary
years
2026 9roles
background 4representative citing papers
PAIR combines a hidden-state probe with an attention correction to deliver robust step-level rewards for GRPO-based optimization of multi-turn LLM agents, achieving high AUROC on contaminated trajectories at low cost.
A learnable continuous perturbation framework for LLM token prefixes via latent vector transformations, optimized through unbiased estimating equations, yields gains in out-of-domain performance.
Perturbing prefixes to semantic neighbors during training creates a hierarchical noise model that improves language model predictions on token sequences outside the training corpus support.
DGPO is a critic-free RL framework that uses bounded Hellinger distance and entropy-gated advantage redistribution to enable fine-grained token-level credit assignment in long CoT generations for LLM alignment, reporting SOTA results on AIME benchmarks.
Kernel smoothing enables accurate low-variance value and gradient estimates for policy optimization in LLM reasoning under tight sampling constraints per prompt.
RTT bridges response-level rubrics to token-level rewards via a relevance discriminator and intra-sample group normalization, yielding higher instruction and rubric accuracy than baselines.
PACEvolve++ uses a phase-adaptive reinforcement learning advisor to decouple hypothesis selection from execution in LLM-driven evolutionary search, delivering faster convergence than prior frameworks on load balancing, recommendation, and protein tasks.
A statistical survey of RLHF for LLM alignment that connects preference learning and policy optimization to models like Bradley-Terry-Luce while reviewing methods, extensions, and open challenges.
citing papers explorer
-
PAIR: Prefix-Aware Internal Reward Model for Multi-Turn Agent Optimization
PAIR combines a hidden-state probe with an attention correction to deliver robust step-level rewards for GRPO-based optimization of multi-turn LLM agents, achieving high AUROC on contaminated trajectories at low cost.