Action Inference By Maximising Evidence: Zero-shot Imitation From Observation With World Models

Abstract

Unlike most reinforcement learning agents which require an unrealistic amount of environment interactions to learn a new behaviour, humans excel at learning quickly by merely observing and imitating others. This ability highly depends on the fact that humans have a model of their own embodiment that allows them to infer the most likely actions that led to the observed behaviour. In this paper, we propose Action Inference by Maximising Evidence (AIME) to replicate this behaviour using world models. AIME consists of two distinct phases. In the first phase, the agent learns a world model from its past experience to understand its own body by maximising the ELBO. While in the second phase, the agent is given some observation-only demonstrations of an expert performing a novel task and tries to imitate the expert's behaviour. AIME achieves this by defining a policy as an inference model and maximising the evidence of the demonstration under the policy and world model. Our method is "zero-sh

Stats

• citations0
• S2 citations—
• github stars13
• HF likes0
• heat score2.29
• arxiv keyzhang2023action

Related papers

• Mimicking Better By Matching The Approximate Action Distribution (2023)0.00
• Consistent Zero-shot Imitation With Contrastive Goal Inference (2025)0.00
• Active Inference And Reinforcement Learning: A Unified Inference On Continuous State And Action Spaces Under Partial Observability (2022)5.84
• Reward Maximisation Through Discrete Active Inference (2020)10.74
• Adversarial Imitation Learning Via Random Search (2020)7.16
• Fully General Online Imitation Learning (2021)0.00
• R-AIF: Solving Sparse-reward Robotic Tasks From Pixels With Active Inference And World Models (2024)4.52
• Learning What To Do By Simulating The Past (2021)0.00