Lighthouse RL is a reinforcement learning method for analog circuit sizing that initializes episodes from high-performing configurations ('lighthouses') discovered during training, steering exploration toward promising regions. The approach claims up to 1.72x faster convergence, 100% success rate versus 0-87% for baselines, and improved generalization over standard RL and Bayesian optimization methods. The reset strategy is designed as a plug-and-play enhancement compatible with any RL-based optimizer. The application domain is analog circuit sizing, a computationally expensive black-box optimization problem.
Researchers introduce Reinforcement Learning with Metacognitive Feedback (RLMF), a training paradigm that refines preference optimization using a model's self-judgments of its own performance quality. The method is applied to faithful calibration — aligning a model's expressed confidence with its intrinsic uncertainty — and achieves state-of-the-art results across diverse tasks while outperforming standard RL by up to 63%. A companion technique, metacognitive data selection, uses similar self-judgments to identify high-value training examples, outperforming naive active learning baselines. The work positions metacognitive performance as a novel and effective RL signal for improving LLM reliability and alignment.
A new arXiv preprint provides theoretical analysis of Reinforcement Learning from Verifiable Rewards (RLVR) updates, identifying off-policy degree and gradient expectation as key factors governing update dynamics. The authors show that differences in gradient steps per rollout substantially affect importance sampling ratio distributions and which tokens dominate updates. Based on this analysis, they propose Adaptive Clip Policy Optimization (ACPO), which adjusts clipping boundaries per token group by empirical variance of importance sampling ratios, outperforming DAPO and CISPO baselines on 3B and 7B models across math, tabular QA, and logic benchmarks.
Hugging Face published a blog post introducing RLOO (REINFORCE Leave-One-Out), a reinforcement learning algorithm aimed at making the RL component of RLHF more practical and effective. The post discusses implementation details and motivations for revisiting pure RL-based fine-tuning approaches within the TRL library. This represents a technical contribution to the alignment and RLHF tooling ecosystem, offering an alternative to PPO-based RLHF pipelines.
Researchers propose PACT (Plan, Align, Commit, Think), a hybrid architecture pairing a fast reactive RL policy with an asynchronous small language model planner for deliberation. The SLM generates and validates candidate action plans via simulation before committing to execution, bypassing the RL policy without retraining. Evaluated on FrozenLake configurations of increasing difficulty, PACT outperforms baselines using only a 2B-parameter SLM, suggesting complementary strengths between deliberative planning and reactive execution.
Hugging Face published a detailed tutorial demonstrating how to fine-tune Meta's LLaMA model using Reinforcement Learning from Human Feedback (RLHF) on StackExchange data. The guide covers the full pipeline: supervised fine-tuning, reward model training, and PPO-based RL optimization. It serves as a practical reference for practitioners seeking to replicate RLHF workflows on open-weight models using the TRL library.
SafeCtrl-RL is a framework for controlling LLM safety at inference time without retraining or modifying model parameters. It formulates dialogue generation as a sequential decision process where an RL agent dynamically selects prompt adjustment strategies based on contextual feedback, iteratively suppressing unsafe outputs. The authors frame this as 'inference-time behavioural unlearning' and report improvements in safety and response quality across multiple LLMs and unsafe dialogue scenarios, outperforming existing prompt-based optimisation baselines.
This arXiv paper proposes Output Reset (OR), a smooth one-sided saturation rule to replace the clipped surrogate objective in PPO and GRPO during LLM post-training. Experiments on Llama-3.2-1B-Instruct with the Anthropic hh-rlhf dataset show PPO-OR achieves a 0.305 higher mean reward-model score than PPO-clip under GAE, while GRPO-OR shows reduced variance but no reward gain at group size G=2. The work identifies a meaningful behavioral difference between the two optimization regimes but leaves open whether larger group sizes change GRPO-OR's effectiveness.
Researchers introduce RePro (Retrospective Progress-Aware Training), a framework addressing the gap between step-wise RL optimization and metacognitive task-progress awareness in LLM agents. The approach uses a forward-then-reflect rollout paradigm where agents execute actions online and then retrospectively assess step-wise progress given the completed trajectory and known outcome. Evaluated on WebShop, ALFWorld, and Sokoban, RePro achieves up to 12% absolute success rate gains over baseline Qwen-family models without requiring continuous external supervision.