The paper presents an online learning algorithm that significantly boosts data efficiency for reinforcement learning from human feedback (RLHF). It incrementally updates reward and language models as choice data arrives, using a small affirmative nudge, an epistemic neural network for reward uncertainty, and information‑directed exploration. With Gemma LLMs, the method matches offline RLHF trained on 200K labels using fewer than 20K labels, achieving over a 10× improvement in data efficiency, and projects a 1,000× gain when scaled to 1M labels.
arXiv:2306. 09712v2 Announce Type: replace Abstract: In reinforcement learning (RL), there are two major settings for interacting with the environment: online and offline.
The paper introduces an exploration-guided prompt scaffolding framework for multimodal large language models, dynamically adjusting the prompt distribution during reinforcement learning post-training. It uses an Exploration Potential Score (EPS) derived from KL-regularized policy improvement to assess prompt utility without extra overhead, and a teacher model rewrites low-utility prompts to preserve intent while improving informativeness. Experiments on Geo3K, MMK12, MathVision, and MMMU-Pro show consistent performance gains, up to 9.7% in-domain and over 11% on out-of-distribution benchmarks.
arXiv:2604. 17244v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents for sequential decision-making struggle to produce diverse outputs.
arXiv:2510. 11686v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) promises to expand the capabilities of language models, but it is unclear if current RL techniques promote the discovery of novel behaviors, or simply sharpen those already present in the base model.
The paper "Demystifying Reinforcement Learning Post-Training of Language Models" investigates how reinforcement learning (RL) post‑training enhances large language models (LLMs) for tasks such as reasoning, math, and coding. By isolating RL components in a controlled setting, the authors analyze how the base model’s prior distribution, reward granularity, prompt diversity, and model scale influence outcomes, using policy entropy to compare pre‑training, supervised fine‑tuning (SFT), and RL stages. The study clarifies the role of spurious rewards, the importance of the base model’s probability mass on desired behaviors, and how these factors interact to determine post‑training success, offering a practical primer for NLP researchers. "whyItMatters":"The work provides a clearer understanding of RL post‑training mechanics, helping researchers and practitioners effectively apply RL to improve LLM capabilities."
arXiv:2608. 16707v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed as decision-making agents in settings that require sophisticated environmental exploration.
arXiv:2502. 19544v3 Announce Type: replace Abstract: Leveraging offline data is a promising way to improve the sample efficiency of online reinforcement learning (RL).
arXiv:2509. 24372v3 Announce Type: replace-cross Abstract: Fine-tuning large language models (LLMs) for downstream tasks is an essential stage of modern AI deployment.
arXiv:2607. 10738v1 Announce Type: cross Abstract: Recent advances in equipping Large Language Models (LLMs) with search tools and outcome-reward reinforcement learning (RL) have achieved new state-of-the-art results on open-domain QA tasks.
arXiv:2509. 02522v3 Announce Type: replace-cross Abstract: Recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have empowered large language models (LLMs) to tackle challenging reasoning tasks such as mathematics and programming, however existing RLVR methods often suffer from sparse reward signals and unstable policy gradient updates inherent to RL-based approaches.
arXiv:2606. 27814v1 Announce Type: new Abstract: Training small language-model agents for long-horizon interactive tasks requires both fast imitation and reward-driven improvement.
arXiv:2606. 27814v4 Announce Type: replace Abstract: Training small language-model agents for long-horizon interactive tasks requires both fast imitation and reward-driven improvement.