Policy optimisation, reward modelling and RLHF — how models are trained by feedback rather than by labels.
arXiv:2609.17180v1 Announce Type: new Abstract: Multimodal counselor response generation (MCRG) aims to generate an appropriate counselor response from multimodal dialogue histories. Progress is limi...
arXiv:2609.17325v1 Announce Type: new Abstract: Biological cells can be viewed as individual, interacting agents whose collective dynamics give rise to adaptive behaviour at multiple levels of organi...
arXiv:2609.17416v1 Announce Type: new Abstract: Voice agents built on LLMs follow a rigid listen-think-speak loop that inserts seconds of dead air before every reply. We show that continuous-time cog...
arXiv:2609.17523v1 Announce Type: new Abstract: We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researcher...
arXiv:2609.17221v1 Announce Type: cross Abstract: Autonomous Software Engineering Agents (SWE-Agents) excel in deterministic coding tasks but struggle with Architecture 0, the nascent system design p...
arXiv:2604.18616v2 Announce Type: replace-cross Abstract: LLM coding agents can generate correct GPU kernels, but their performance still trails expert libraries. Reaching peak throughput requires co...
arXiv:2609.13083v2 Announce Type: replace-cross Abstract: In urban search and rescue, articulated tracked robots (ATRs) must traverse structured but contact-rich environments such as stairwells and c...
arXiv:2609.16059v1 Announce Type: cross Abstract: Multimodal instruction following (MMIF) is crucial for building generalist agents. However, current training paradigms rely heavily on Supervised Fin...
arXiv:2609.16660v1 Announce Type: new Abstract: Test-time reinforcement learning adapts a model on its own unlabeled test set using majority-vote pseudo-labels and has shown strong results in mathema...
The paper introduces JudgeBiasBench, a benchmark that systematically quantifies judgment biases in large language model (LLM)-based judges across four dimensions and 12 bias types. It evaluates both generative and discriminative judges, revealing significant bias patterns that undermine reliability. The authors propose bias-aware training—reinforcement learning for generative judges and contrastive learning for discriminative judges—to reduce these biases while maintaining evaluation performance.
We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyday workflows. ScienceBuddy supports rese...
Decentralized bandit systems often contain heterogeneous agents: rewards can change at individual agents even when the best action for the network stays the same. These local changes may cancel when r...
With the rapid development of Large language model (LLM), agent systems enhanced by LLMs show huge potential in being able to deal with complex tasks, especially involving multi-step thinking or inter...
We give deterministic and uncoupled learning dynamics for finite multiplayer general-sum games under full-information feedback that achieve constant individual swap regret, independent of the horizon...
Automated depression screening from clinical interviews requires attribution of utterances to the clinician or patient. We evaluate two datasets: DAIC-WOZ, where participant-only recordings require re...
arXiv:2609.13676v1 Announce Type: new Abstract: Markov decision processes (MDPs) are used to support decision-making in conservation of biodiversity, but policies, even over small state spaces, can b...
arXiv:2609.15915v1 Announce Type: new Abstract: Meta-reinforcement learning (meta-RL) enables agents to adapt to unseen tasks with limited experience. Despite its promise, the application of meta-RL...
Zonal RL-RRT is a new path‑planning algorithm that partitions a map into zones using kd‑tree partitioning and employs Value Iteration as a high‑level decision maker. The method achieves a three‑fold improvement in time efficiency over basic sampling methods such as RRT and RRT* in forest‑like maps, and outperforms heuristic‑guided methods like BIT* and Informed RRT* by 1.5× in runtime while maintaining robust success rates across 2D to 6D environments. It also shows on average a 1.5× better performance than learning‑based methods such as NeuralRRT* and MPNetSMP, and has been validated in simulations of a UR10e arm manipulator in MuJoCo.
CHAI for LLMs is a framework that improves large language models’ performance on code‑mixed translation tasks by using LLMs as annotators to create preference data, applying reinforcement learning from AI feedback, incorporating LLM‑generated domain knowledge for iterative refinement, and evaluating on real‑world datasets. The approach yields a 68.45% average win rate over state‑of‑the‑art open‑source models in human‑adjudicated tests. It demonstrates a scalable method to enhance code‑mixed language understanding in open‑source LLMs.
arXiv:2609. 13954v1 Announce Type: new Abstract: Ensemble sampling offers a practical approach to randomized exploration by maintaining a collection of models, but how small an ensemble can be while retaining strong regret guarantees remains unresolved.