Policy optimisation, reward modelling and RLHF — how models are trained by feedback rather than by labels.
arXiv:2405.08253v4 Announce Type: replace-cross Abstract: This paper develops a framework for learning in discounted infinite-horizon Markov decision processes (MDPs) with Borel state and action spac...
arXiv:2604.01345v3 Announce Type: replace Abstract: Inverse reinforcement learning (IRL) recovers the loss function of a forward learner from its observed responses. Adaptive IRL aims to reconstruct...
arXiv:2511.07260v3 Announce Type: replace-cross Abstract: Ad hoc teamwork (AHT) requires agents to collaborate with previously unseen teammates, which is crucial for many real-world applications. The...
arXiv:2601.13519v4 Announce Type: replace-cross Abstract: This paper introduces a new problem-dependent regret measure for online convex optimization with smooth losses. The notion, which we call the...
arXiv:2609.16751v2 Announce Type: replace-cross Abstract: We give deterministic and uncoupled learning dynamics for finite multiplayer general-sum games under full-information feedback that achieve c...
arXiv:2609.22221v1 Announce Type: new Abstract: Large language models (LLMs) can generate fluent and coherent text that is increasingly difficult to distinguish from human writing, motivating the dev...
arXiv:2609.24194v1 Announce Type: new Abstract: Evaluation scores used around LLM systems -- including reward models, rerankers, and LLM judges -- can track surface form instead of the quality they c...
arXiv:2605.25920v2 Announce Type: replace Abstract: While large language models (LLMs) augmented with agentic search capabilities show promise for legal reasoning, they overlook a fundamental constra...
arXiv:2606.22570v2 Announce Type: replace Abstract: Reinforcement Learning from Verifiable Rewards (RLVR) has emerged as a promising framework for enhancing the reasoning ability of large language mo...
arXiv:2601.12538v2 Announce Type: replace-cross Abstract: Reasoning is a fundamental cognitive process underlying inference, problem-solving, and decision-making. While large language models (LLMs) d...
arXiv:2609.24489v1 Announce Type: new Abstract: Offline reinforcement learning (RL) typically trains a critic by minimizing a regression loss against bootstrapped value targets stabilized by target n...
arXiv:2609.24526v1 Announce Type: new Abstract: Physical AI requires models to ground visual and linguistic understanding in real-world environments while accounting for environmental constraints and...
Contrastive World Models propose a new method for learning latent dynamics without pixel reconstruction. By replacing observation reconstruction with a Deep InfoMax-like objective that maximizes mutual information between state-action sequences and local patch features of future observations, the approach encourages state representations to retain predictive information while ignoring visually irrelevant details. Experiments show that this method matches existing baselines in simple settings and significantly outperforms them when distractors or natural video backgrounds are present, while also training more efficiently by eliminating the pixel decoder.
The study investigates whether learned scoring functions can outperform hand‑tuned heuristics in Kubernetes node‑selection. Two models—a Random Forest on engineered features and a graph neural network on job dependency graphs—were trained on a large production trace; both achieved modest regression gains (R²≈0.042) but lagged behind a simple free‑CPU heuristic in Top‑1 ranking accuracy (65‑66% vs. 74‑84%). The authors attribute this gap to objective mismatch, noting that pointwise regression rather than a ranking‑specific loss likely limits performance.
The paper investigates using offline reinforcement learning to improve line‑selective tripping in distribution grids. A convolutional Q‑network trained with conservative Q‑learning (CQL) processes voltage‑current phasor and impedance data, optionally with raw waveforms, to predict faulted lines. On a realistic CIGRE medium‑voltage network, the best model achieved high per‑timestep precision, recall, and F1‑score, and correctly identified the first trip action in over 98% of fault episodes, though it mis‑tripped in a notable fraction of non‑fault cases.
The paper reviews the evolution of multi‑agent unmanned systems from isolated sensing to collaborative intelligence, where agents share compact features to overcome local observation limits such as occlusions and sensor range. It introduces a five‑dimensional taxonomy (collaboration stage, communication paradigm, fusion architecture, learning strategy, application domain) and three cognitive synergy conditions (Semantic Disambiguation, Pragmatic Information Exchange, Proactive Informational Foraging) to unify existing research. The authors survey architectures, neural‑communication co‑design, embodied action‑perception loops, and resilience mechanisms, map advances onto operational domains (V2X, UAV, logistics, smart cities), and propose the GCI‑Bench scoring protocol to standardize evaluation across studies.
arXiv:2609. 22833v1 Announce Type: new Abstract: We study personalized federated reinforcement learning, in which $n$ agents, each acting in its own Markov decision process, collaborate through a server to learn a shared MAML-style policy initialization that becomes effective for an individual agent once that agent adapts it with a single local policy-gradient step.
The paper introduces Calibrated Clipping, a dynamic method to align FP8 quantization bounds with high‑precision BF16 distributions, thereby mitigating training instability in full‑pipeline FP8 reinforcement learning for large language models. It identifies that compounded FP8 noise distorts importance ratios, causing entropy surges and garbled outputs. Experiments across GRPO and DAPO algorithms on 8B‑32B models show the technique restores performance to BF16 levels.
The paper introduces CNA, an AI‑oriented comprehensive normalized assessment that transforms vital‑sign distributions into a standard normal space to produce a unified health‑status score. This score serves as a criterion for evaluating and terminating reinforcement learning policies in Renal Replacement Therapy (RRT). Using a 23‑dimensional state representation and matrix decomposition to handle missing data, the authors apply offline RL algorithms and show that the learned strategy reduces mortality from 13.2% to 5.0% and shortens hospital stays by nearly 59 hours compared to physician‑observed treatments.
The paper proposes a hybrid PID–Deep Reinforcement Learning (DRL) controller for industrial processes, addressing the limitations of traditional PID controllers in complex, non‑linear, multi‑input environments. Using the Industrial Benchmark (IB) to test DRL, the authors develop a multi‑objective reward function and employ a TD3 agent to discover optimal settings for the IB’s ‘Gain’ and ‘Shift’ parameters. These parameters are then fed into a tuned PID controller, yielding a system that combines the optimal performance and efficiency of DRL with the reliability of classical control.