AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Revisiting Regularized Policy Optimization for Stable and Efficient Reinforcement Learning in Two-Player Games

Kazuki Ota, Takayuki Osa, Motoki Omura, Tatsuya Harada

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Model-based RL with guarantees · 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
KLENT instead of Gumbel AlphaZero for learning efficiencyGumbel AlphaZero · Five board games averagesimulator evaluations to 50% win rate: 300 million -> 75 millionImproved
KLENT with test-time MCTS versus Gumbel AlphaZeroGumbel AlphaZero · Five board games average (test-time MCTS)win rate (%): 53.6% -> 77.2%Improved
KLENT versus Gumbel AlphaZero on 9x9 GoGumbel AlphaZero · 9x9 Go (test-time MCTS)win rate (%): 37% -> 89%Improved
KLENT versus Gumbel AlphaZero on HexGumbel AlphaZero · Hex (test-time MCTS)win rate (%): 47% -> 98%Improved
KLENT versus Gumbel AlphaZero on Animal ShogiGumbel AlphaZero · Animal Shogi (test-time MCTS)win rate (%): 67% -> 63%Hurt
KLENT versus Gumbel AlphaZero on OthelloGumbel AlphaZero · Othello (test-time MCTS)win rate (%): 47% -> 55%Improved
add reverse-KL regularization (vs entropy only) (ablation)KLENT · Five board gameswin rate learning curveImproved
add entropy regularization (vs KL only) (ablation)KLENT · Five board gameswin rate learning curveImproved
use lambda-returns instead of 1-step returns (ablation)KLENT · 9x9 Go and Hexwin rate learning curveImproved
use lambda-returns instead of Monte Carlo returns (ablation)KLENT · 9x9 Go and Hexwin rate learning curveImproved
KLENT versus AlphaZero on 19x19 GoAlphaZero · 19x19 Golearning curveNo clear effect
KLENT versus AlphaZero on Gardner ChessAlphaZero · Gardner Chess (test-time MCTS)win rate (%): 64% -> 81%Improved

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