AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

Adversarial Imitation Learning With General Function Approximation: Theoretical Analysis and Practical Algorithms

Tian Xu, Zhilong Zhang, Zexuan Chen, Ruishuo Chen, Yihao Sun, Yang Yu

TPAMI 2026 · Reinforcement learning › Imitation learning · approach: Adversarial imitation learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add optimism regularization to policy learning (ablation)MF OPT-AIL and MB OPT-AIL · DMControl 8 tasks, 1 expert demolearning curve convergenceImproved
use MF OPT-AIL instead of HyPEHyPE · DMControl average, 1 demopolicy return: 534.0 -> 683.0Improved
use MB OPT-AIL instead of HyPERHyPER · DMControl average, 1 demopolicy return: 613.4 -> 691.4Improved
use MF OPT-AIL instead of BCBC · DMControl average, 1 demopolicy return: 82.0 -> 683.0Improved
increase expert demos from 1 to 10MF OPT-AIL · DMControl averagepolicy return: 683.0 -> 755.6Improved
increase expert demos from 1 to 10MB OPT-AIL · DMControl averagepolicy return: 691.4 -> 700.0No clear effect
increase expert demos from 1 to 10CMIL · DMControl averagepolicy return: 389.7 -> 450.2Mixed
use OPT-AIL reward update instead of FILTERFILTER · DMControl averagePearson correlation: 0.5868 -> 0.7763Improved
use MB OPT-AIL reward update instead of CMILCMIL · DMControl averagePearson correlation: 0.5748 -> 0.7472Improved
use model-based rather than model-free learningOPT-AIL · DMControl 1 expert demointeraction efficiencyImproved
use MB OPT-AIL on Cheetah Run instead of MF OPT-AILMF OPT-AIL · DMControl Cheetah Run, 1 demopolicy return: 348.7 -> 577.0Improved
use MB OPT-AIL instead of MF OPT-AILMF OPT-AIL · DMControl average, 10 demospolicy return: 755.6 -> 700.0Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy