Policy optimisation, reward modelling and RLHF — how models are trained by feedback rather than by labels.
arXiv:2607. 09641v1 Announce Type: cross Abstract: Financial anomaly detection suffers from extreme class imbalance, causing traditional single-objective algorithms to exhibit ``fraud collapse'', defaulting to the majority class and failing to balance anomaly interdiction with customer friction.
arXiv:2607. 09298v1 Announce Type: cross Abstract: We study general-utility Markov decision processes (GUMDPs) with risk-aware objectives.
arXiv:2607. 08793v1 Announce Type: cross Abstract: Sepsis is a leading cause of mortality, yet optimal treatment policies remain contested.
arXiv:2607. 08971v1 Announce Type: new Abstract: The stochastic linear bandit, where actions are represented as vectors and rewards are linear, is a central paradigm for sequential decision making.
arXiv:2607. 09153v1 Announce Type: new Abstract: Process Reward Models (PRMs) have been proven to be highly effective in guiding test-time scaling (TTS) methods, which significantly boost the capabilities of LLM-based multi-agent systems.
arXiv:2607. 08837v1 Announce Type: cross Abstract: Exploration is essential to RL since a policy cannot improve by repeatedly sampling the behaviors it already prefers.
arXiv:2607. 08877v1 Announce Type: cross Abstract: Pretrained generative robot policies based on flow matching and diffusion have achieved impressive results across a wide range of manipulation tasks.
arXiv:2607. 09422v1 Announce Type: new Abstract: Cooperative multi-agent reinforcement learning is well suited to problems with large parameter spaces and exploitable local structure, such as the tuning of electrostatically-defined quantum-dot arrays.
arXiv:2607. 09042v1 Announce Type: new Abstract: Reinforcement learning (RL) is increasingly used to post-train vision-language-action (VLA) models, but every update consumes robot rollouts that are slow and costly to collect, making sample efficiency a central concern.
Reinforcement learning for large language models (LLMs) typically relies on trust-region masks to stabilize off-policy updates. The dominant PPO-style approach uses the sampled-token importance ratio for two criteria: a proximity criterion, which asks whether the policy has moved too far from the behavior policy, and a direction criterion, which asks whether the update pushes it farther away.
Synthetic data is widely used to train large language models because it is inexpensive to generate and easy to control. As models are increasingly deployed as agents, synthetic trajectories are likely to become an important source of training data for agentic behavior.
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. However, we argue that current training paradigms harbor a critical vulnerability: they predominantly reward correct answers but fail to penalize fabricated ones when retrieval fails, thereby implicitly exacerbating hallucinations.
Agentic retrieval-augmented generation (RAG) extends static RAG by allowing language models to iteratively reason, generate search queries, retrieve evidence, and predict answers. However, it remains challenging for models to decide when to retrieve, whether to use lexical matching or semantic similarity, and how to control context granularity to prevent irrelevant tokens from interfering with agent reasoning.
arXiv:2605. 00155v3 Announce Type: replace Abstract: Reinforcement learning from human feedback (RLHF) is a central post-training tool for aligning large language models, but its training reward is only a learned proxy for true human utility.
arXiv:2607. 07748v1 Announce Type: new Abstract: Large Language Models achieve strong code generation for high resource languages like Python and Java but suffer sharp performance drops on Low-Resource Programming Languages~(LRPLs) such as Julia.
arXiv:2607. 08404v1 Announce Type: cross Abstract: Current computational approaches for drug design typically focus on generating molecules conditioned on specific targets or general molecular properties, often neglecting the influence of disease context on target behavior and therapeutic outcomes.
arXiv:2607. 08647v1 Announce Type: cross Abstract: As autonomous agents are increasingly deployed across diverse operational contexts, aligning their behavior with human intent demands reward functions that remain robust to such changes rather than overfitting to any single environment.
arXiv:2607. 08193v1 Announce Type: cross Abstract: Open-ended curricula in Reinforcement Learning (RL) aim to train generally-capable agents by identifying tasks that facilitate learning increasingly complex skills.
arXiv:2607. 08373v1 Announce Type: cross Abstract: Connected vehicles are autonomous cyber-physical systems whose behavior must be continuously monitored during operation to detect deviations from normal operation before they propagate into failures.
arXiv:2607. 07758v1 Announce Type: new Abstract: Foundation models (FMs) have transformed machine learning from isolated task-specific model development toward general-purpose models pretrained on broad data and adapted to multiple downstream tasks.