Online Experiential Learning for Language Models

TL;DR

This paper introduces Online Experiential Learning (OEL), a framework enabling language models to learn continuously from their deployment experiences, improving performance through iterative knowledge extraction and consolidation without requiring environment access.

Contribution

The paper proposes a novel online learning framework for language models that leverages real-world deployment experiences to enhance performance iteratively.

Findings

OEL improves task accuracy and token efficiency over iterations.

Extracted experiential knowledge outperforms raw trajectories in effectiveness.

On-policy consistency is crucial for effective experiential learning.

Abstract

The prevailing paradigm for improving large language models relies on offline training with human annotations or simulated environments, leaving the rich experience accumulated during real-world deployment entirely unexploited. We propose Online Experiential Learning (OEL), a framework that enables language models to continuously improve from their own deployment experience. OEL operates in two stages: first, transferable experiential knowledge is extracted and accumulated from interaction trajectories collected on the user side; second, this knowledge is consolidated into model parameters via on-policy context distillation, requiring no access to the user-side environment. The two stages are iterated to form an online learning loop, where the improved model collects higher-quality trajectories that yield richer experiential knowledge for subsequent rounds. We evaluate OEL on text-based game environments across multiple model scales and both thinking and non-thinking variants. OEL achieves consistent improvements over successive iterations, enhancing both task accuracy and token efficiency while preserving out-of-distribution performance. Our analysis further shows that extracted experiential knowledge is significantly more effective than raw trajectories, and that on-policy consistency between the knowledge source and the policy model is critical for effective learning.