Variance-Averse n-Step Offline Reinforcement Learning for Sparse Long-Horizon Environments

arXiv:2610.07899v1 Announce Type: cross Abstract: Generative actors are transforming offline reinforcement learning (RL) by enabling expressive policy classes that model complex action distributions. However, this expressiveness also exposes a key challenge in heterogeneous datasets: generative policies can reproduce unreliable action modes whose return distributions exhibit high variance, occasionally yielding high returns by chance but lacking consistency. Consequently, maximizing the expected Q-value alone is insufficient for identifying reliable actions. We propose VAN-Flow (Variance-Averse n-step Flow), a framework that promotes reliable actions in generative offline RL. VAN-Flow combines (i) a categorical distributional critic, (ii) a variance-averse expectation operator that smoothly reweights atom probabilities to favor actions with both high returns and low dispersion, and (iii) a flow-matching generative actor guided via rejection sampling. Unlike CVaR or mean-variance objectives, the operator redistributes probability mass over the categorical return distribution without hard truncation or auxiliary penalty terms. Across more than 40 tasks from D4RL and OGBench, VAN-Flow consistently outperforms strong baselines, with the largest gains in long-horizon and high-variance regimes where reliable action selection becomes critical.

Related

• Offline Reinforcement Learning with Universal Horizon Models
• Generative OOD-regularized Model-based Policy Optimization
• Reinforcement Learning with Action Chunking
• T2SPO: Trajectory-to-Step Policy Optimization for Agentic Reinforcement Learning

Source: arXiv cs.AI | 2026-10-07