arXiv:2606. 30335v1 Announce Type: new Abstract: Autonomous scientific discovery systems increasingly use large language models (LLMs) to propose new hypotheses, but many such systems condition primarily on experimental memory: archives of high-scoring candidates or heuristic summaries of recent trials.
arXiv:2608. 09805v1 Announce Type: cross Abstract: Exploration has been a focus of reinforcement learning research for a long time.
arXiv:2608. 09335v1 Announce Type: new Abstract: Multistage stochastic model predictive control (MPC) handles uncertainty by optimizing over a scenario tree, a finite branching approximation of future outcomes constructed from sampled forecasts.
Latent Energy Action Planning (LEAP) is a new method that treats the entire action horizon as a differentiable variable and optimizes it using a frozen LeWorldModel (LeWM). LEAP couples terminal latent goal matching with a terminal‑window state energy, ensuring that the predicted terminal latent and decoder‑predicted terminal descriptor align with the goal. Using a frozen goal‑conditioned proposal, a quasi‑Newton solver, and post‑optimization projection, LEAP improves mean success from 77.5% to 94.8% across four control domains while keeping the LeWM representation frozen.
The paper introduces EXPO, an algorithm that automatically optimizes the meta-prompt—specifically the task description and meta-instruction—for large language model agents in sequential decision-making tasks such as Bayesian optimization and multi-armed bandits. Building on adversarial bandit techniques to handle non-stationary rewards, the authors extend EXPO to EXPO-ES, which also optimizes exemplars (historical interactions) within the meta-prompt. Experiments demonstrate that these methods significantly improve the performance of LLM-based agents in sequential decision-making scenarios.
arXiv:2610.02198v1 Announce Type: cross Abstract: Several state-of-the-art methods for online reinforcement learning in continuous control improve policies using action gradients of a learned critic....
TIGPO (Temporal Instance-Graph Policy Optimization) extends graph-based credit assignment for long-horizon LLM agents by maintaining a persistent transition graph per task across policy updates. It allocates rollout budgets to both new exploration and revisiting past tasks, pairing current rollouts with earlier ones to create cross‑temporal references that stabilize advantage estimation. Experiments on ALFWorld and WebShop show TIGPO consistently outperforms previous group‑based and graph‑based policy optimization methods.
arXiv:2606. 19236v1 Announce Type: cross Abstract: Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training.
Long-horizon large language model (LLM) agents are typically optimized with sparse terminal outcomes, making fine-grained credit assignment across multi-step interactions difficult. Existing approache...
arXiv:2607. 17973v1 Announce Type: new Abstract: Latent world models have emerged as a powerful planning paradigm by learning action-conditioned predictive dynamics and using them as internal simulators to imagine and evaluate candidate action sequences.
arXiv:2604. 18701v3 Announce Type: replace-cross Abstract: Local prediction-error-based curiosity rewards focus on the current transition without considering the world model's cumulative prediction error across all visited transitions.
arXiv:2608. 19842v1 Announce Type: new Abstract: Agentic reinforcement learning (RL) has become a critical stage in the post-training of large language models.