Policy optimisation, reward modelling and RLHF — how models are trained by feedback rather than by labels.
arXiv:2609.24290v1 Announce Type: new Abstract: Instruction-tuned LLMs faced with underspecified queries often commit to a single interpretation rather than ask for clarification, producing confident...
arXiv:2609.23665v1 Announce Type: cross Abstract: Constraint-following music generation asks a score to satisfy several user-specified properties at once, each checkable programmatically (key, meter,...
arXiv:2609.24456v1 Announce Type: cross Abstract: Distributed reinforcement learning (RL) scales training by parallelizing actors and learners around an Experience Buffer. As RL workloads grow, howev...
arXiv:2603.02266v2 Announce Type: replace-cross Abstract: Test-Time Scaling has shown notable efficacy in addressing complex problems through scaling inference compute. However, within Large Audio-La...
arXiv:2603.03768v2 Announce Type: replace-cross Abstract: Full-stack human-robot collaboration (HRC) can become brittle when replacing a planner, partner model, coordination policy, or controller cha...
arXiv:2609.25645v1 Announce Type: new Abstract: Exhaustively evaluating every candidate LLM configuration on every benchmark item to identify a high-performing one is costly. We formulate configurati...
arXiv:2609.26275v1 Announce Type: new Abstract: The vehicle routing problems with real-world constraints (we consider vehicles capacity limits, time windows constrains, pickup-and-delivery multi-depo...
arXiv:2609.26355v1 Announce Type: new Abstract: Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted ma...
arXiv:2510.10730v3 Announce Type: replace Abstract: We provide a unified algorithmic framework for ensemble sampling in nonlinear contextual bandits and develop corresponding regret bounds for two mo...
arXiv:2609.14711v2 Announce Type: replace Abstract: Bayesian quantum tomography requires efficient inference while preserving a posterior fixed by the prior and Born likelihood. Learned transport pro...
arXiv:2609.25270v1 Announce Type: new Abstract: Generating Scalable Vector Graphics (SVG) code from natural-language instructions is an open-ended task without absolute visual ground truth, leaving b...
arXiv:2609.25831v1 Announce Type: cross Abstract: Recent VLM-based autonomous driving planners adopt GRPO-style reinforcement learning to optimize driving performance. However, existing GRPO recipes...
arXiv:2609.26213v1 Announce Type: cross Abstract: We study the multiplayer multi-armed bandit problem with information asymmetry under Bernoulli rewards, for three information structures: asymmetry i...
The paper introduces a method for creating reactive character behaviors in continuous games as compact, human‑readable programs. It searches over a domain‑specific language that uses reactive geometric decisions and higher‑order constructs to discretize continuous behavior space, while eliminating redundant program forms through synthesis antipatterns. The approach, called agentic sketching, combines bottom‑up symbolic enumeration with top‑down guidance from a coding agent, and outperforms either technique alone on a benchmark of 14 continuous games.
The paper proposes a method for allocating a limited budget of expert annotations to optimize the accuracy of off-policy evaluation in settings where rewards are missing or noisy. By deriving variance‑optimal annotation probabilities for sequential, forward‑monotone protocols, the authors provide a batch‑adaptive implementation that can be applied to real data. Experiments on casenotes from a homelessness services nonprofit and on human‑preference votes from LMArena demonstrate substantial reductions in RMSE—up to 65% for housing placement and 68% for progress toward a housing application—when using only 40% or more of the full annotation budget.
Agent0 is a fully autonomous framework that enables large language model agents to evolve without external data by using a multi‑step co‑evolution process. It pits a curriculum agent against an executor agent, both derived from the same base LLM, where the curriculum agent creates increasingly challenging tasks and the executor learns to solve them. By integrating external tools into the executor’s workflow, the system creates a self‑reinforcing cycle that continuously generates high‑quality curricula, leading to significant gains in reasoning performance—an 18% improvement on mathematical reasoning and 24% on general reasoning for the Qwen3‑8B‑Base model.
The paper presents a spiking neural network-based Proximal Policy Optimization (SNN‑PPO) algorithm for autonomous UAV navigation through constrained openings in civil infrastructure. By integrating spike‑based actor‑critic reinforcement learning with a stochastic Gaussian policy, the method achieved 63.77 % overall success across 3000+ episodes, improving to over 90 % in later stages and averaging 2.10 windows per episode.
The paper introduces a decentralized decision-making framework for teams operating under partial observability and unknown system dynamics. By leveraging low-rank latent dynamics and delayed shared information, each team member learns an approximate Markov decision process using only local private data and delayed common updates. The resulting algorithm achieves near‑optimal team performance without requiring a centralized coordinator or training, and the authors provide finite‑sample guarantees and a sample‑complexity bound.
The paper investigates best‑policy identification in finite‑horizon, risk‑sensitive reinforcement learning using the entropic risk measure. It identifies a gap between known lower bounds ≥ η(e^{|eta|H}) and upper bounds ≤ O(e^{2|eta|H}) for sample complexity, attributing the excess factor to loose concentration bounds for exponential utilities. By employing a forward‑model algorithm with KL‑based exploration bonuses and a novel stopping rule, the authors achieve a sample complexity that matches the lower bound, closing the previously open exponential gap.
arXiv:2609.24101v1 Announce Type: new Abstract: Automated biomedical evidence synthesis depends on retrieving published studies, but the biomedical literature is systematically skewed toward positive...