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.
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. To build such a tree, conventional methods focus on matching the underlying probability distribution---e.
arXiv:2407. 21359v2 Announce Type: replace-cross Abstract: Imagining potential outcomes of actions before execution helps agents make more informed decisions, a prospective thinking ability fundamental to human cognition.
arXiv:2607. 09298v1 Announce Type: cross Abstract: We study general-utility Markov decision processes (GUMDPs) with risk-aware objectives.
arXiv:2605. 04568v3 Announce Type: replace-cross Abstract: State-of-the-art model-based Reinforcement Learning (RL) approaches either use gradient-free, population-based methods for planning, learned policy networks, or a combination of policy networks and planning.
arXiv:2608. 02519v1 Announce Type: new Abstract: Effective model-based reinforcement learning in stochastic environments requires planning that accounts for predictive uncertainty.
arXiv:2607. 26509v1 Announce Type: new Abstract: Deep off-policy reinforcement learning algorithms for continuous control typically rely on neural value function approximation to guide policy improvement.
OneBid is a unified auto‑bidding foundation model that consolidates diverse cost‑per‑X (oCPX) advertising scenarios into a single framework. It builds on Decision Transformer by conditioning on two atomic signals—Return‑to‑Go for conversion value and Cost‑to‑Go for cost ratio—and incorporates value‑aware regularization. A sequence‑level Mixture‑of‑Experts architecture captures cross‑scenario knowledge while preserving low latency, and a Critic‑guided Relative Offline Policy optimization (CROP) aligns the backbone with scenario‑specific preferences without unsafe online exploration. In production at Kuaishou, OneBid achieved a 2.2% overall ADVV increase and up to 13.1% in the ROAS scenario.
EPIG-Tree proposes a compute‑optimal branching strategy for gradient‑efficient reinforcement learning, arguing that branches should be placed where they most reduce policy‑gradient uncertainty per unit of compute. By deriving allocation laws from a law‑of‑total‑variance decomposition, the method introduces an EPIG‑Tree score that guides branch placement using already computed rollouts, estimating occupancy‑ and score‑weighted value uncertainty. Empirical results show EPIG‑Tree reduces gradient MSE in cloned‑state control, improves frozen‑LLM gradient calibration, and outperforms flat GRPO and entropy branching in both single‑turn math and multi‑turn Wordle tasks.
arXiv:2609.40035v1 Announce Type: new Abstract: The policy-gradient theorem gives the exact gradient under the current policy, but finite on-policy samples may miss rare high-return trajectories. We...
BCPPO is a new variant of Proximal Policy Optimization that uses Bachelier-inspired cost‑prediction networks to generate a smooth penalty based on disagreement among critics. The method keeps temporal‑difference learning unchanged, applies a saturation‑aware controller to manage cost penalties, and deploys only the policy network. Across extensive experiments, BCPPO outperforms comparators in achieving higher mean returns while maintaining lower or comparable CVaR in all tested tasks.
Effective model-based reinforcement learning in stochastic environments requires planning that accounts for predictive uncertainty. Propagating full state distributions analytically offers a principled way to do this, but has traditionally required restrictive policy or reward structures to remain tractable.
arXiv:2608. 19842v1 Announce Type: new Abstract: Agentic reinforcement learning (RL) has become a critical stage in the post-training of large language models.