When Does On-Policy Interaction Help? Representational Tradeoffs in Value-Based Imitation Learning

2026-07-31

AI Summary

This work addresses the performance degradation and error accumulation in behavioral cloning caused by insufficient policy representational capacity. To overcome this limitation, the paper proposes a value-function-based online imitation learning framework that shifts the learning objective from directly mimicking the expert policy to approximating the expert’s value function through online policy interaction. This reformulation substantially reduces the reliance on the learner’s policy expressivity. Theoretical analysis demonstrates that online interaction eliminates dependence on policy class complexity and yields convergence guarantees requiring only the realizability of the value function. Empirical results show that the proposed Online Value Imitation (OVI) algorithm significantly outperforms baseline methods such as Behavioral Cloning (BC) and DAgger, particularly when the learner’s network capacity is substantially weaker than that of the expert.

Abstract

Imitation learning (IL)---training an agent to replicate expert behavior from demonstrations---underpins applications from robotics to language model training. Standard approaches such as Behavior Cloning (BC) are known to suffer from compounding errors and performance plateaus, particularly when the learner cannot perfectly represent the expert's policy (as is typical, e.g., in distillation). Two interventions are widely understood empirically to improve performance: querying the expert interactively along the learner's own trajectories, and using value function estimation en route to generating a policy rather than directly fitting the expert's full action distribution. We investigate the nature of these improvements and their potentially surprising interplay. Our main finding is that expert interaction relaxes the representational demands on the learner: one only needs a model capable of realizing the expert's value function, bypassing the (often stricter) requirement of realizing the expert's policy itself. Concretely, we introduce OVI, an interactive on-policy IL algorithm that is statistically efficient whenever the learner can represent the expert's value function and computationally efficient given access to a linear maximization oracle. We complement this with a negative result showing that interaction is necessary. Namely, without stronger assumptions beyond expert-value realizability alone, any offline IL algorithm must scale with the complexity of the expert policy class. Our findings bear out empirically. OVI outperforms offline policy-based (BC), interactive policy-based (DAgger), and offline value-based IL methods, with the largest gains when the learner network is substantially less expressive than the expert's.

Problem

• Imitation Learning
• Value Function
• Policy Representation
• On-Policy Interaction
• Representational Complexity

Innovation

• on-policy interaction
• value-based imitation learning
• representational efficiency
• OVI algorithm
• expert querying