Research map / Optimization / Nonconvex and smooth optimization

Efficient Last-Iterate Convergence in Solving Extensive-Form Games

NeurIPS 2025 · Optimization › Nonconvex and smooth optimization · approach: Equilibrium learning in zero-sum games · 0 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Studies efficient last-iterate convergence for CFR-type algorithms in solving extensive-form games by reducing the original game to a sequence of perturbed regularized EFGs.

Problem

Last-iterate convergence of counterfactual regret minimization (CFR) algorithms for learning Nash equilibria in extensive-form games, which prior work established via sequences of perturbed regularized EFGs.

Approach

Reformulates learning the NE of the original EFG as learning NEs of a sequence of perturbed regularized EFGs, so last-iterate convergence of the original reduces to that of the regularized subproblems; the paper focuses on making this convergence efficient. Details beyond the abstract are not available.

Experimental setup

Builds on: counterfactual regret minimization

Closest papers by meaning

• The Power of Regularization in Solving Extensive-Form Games (ICLR 2023 · similarity 0.94)
• Last-iterate Convergence in Extensive-Form Games (NeurIPS 2021 · similarity 0.93)
• From Average-Iterate to Last-Iterate Convergence in Games: A Reduction and Its Applications (NeurIPS 2025 · similarity 0.90)
• Fast Last-Iterate Convergence of Learning in Games Requires Forgetful Algorithms (NeurIPS 2024 · similarity 0.90)
• Last-Iterate Convergence Properties of Regret-Matching Algorithms in Games (ICLR 2025 · similarity 0.89)
• Near-Optimal Last-Iterate Convergence for Zero-Sum Games with Bandit Feedback and Opponent Actions (NeurIPS 2026 · similarity 0.88)