Intrinsic Credit Assignment for Long Horizon Interaction

2026-02-12

Technology Category

Humans and AI: Human-Aware Planning and Behavior PredictionMultiagent Systems: Adversarial AgentsMachine Learning: Online Learning & Bandits

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingUser Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

How can we train agents to navigate uncertainty over long horizons? In this work, we propose {\Delta}Belief-RL, which leverages a language model's own intrinsic beliefs to reward intermediate progress. Our method utilizes the change in the probability an agent assigns to the target solution for credit assignment. By training on synthetic interaction data, {\Delta}Belief-RL teaches information-seeking capabilities that consistently outperform purely outcome-based rewards for Reinforcement Learning, with improvements generalizing to out-of-distribution applications ranging from customer service to personalization. Notably, the performance continues to improve as we scale test-time interactions beyond the training horizon, with interaction-efficiency increasing even on Pass@k metrics. Overall, our work introduces a scalable training strategy for navigating uncertainty over a long-horizon, by enabling credit assignment to intermediate actions via intrinsic {\Delta}Belief rewards.

Problem

• credit assignment
• long-horizon interaction
• uncertainty navigation
• reinforcement learning
• intermediate rewards

Innovation

• ΔBelief-RL
• intrinsic credit assignment
• long-horizon reinforcement learning
• language model beliefs
• information-seeking behavior