# Abstraction Agent

arXiv: https://arxiv.org/abs/2609.04303 · HTML: https://mlsanity.com/paper/2609.04303/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use.
- Builds on 25 of its 54 references that are in the map.
- First posted on arXiv 3 Sep 2026.

## Placed in

- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 25 of 25)

- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [DeepStack: Expert-Level Artificial Intelligence in No-Limit Poker](https://mlsanity.com/paper/1701.01724/) (2017)
- [Large Language Models for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineering](https://mlsanity.com/paper/2305.03403/) (2023)
- [Deep Counterfactual Regret Minimization](https://mlsanity.com/paper/1811.00164/) (2018)
- [Suspicion-Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4](https://mlsanity.com/paper/2309.17277/) (2023)
- [Strategic Reasoning with Language Models](https://mlsanity.com/paper/2305.19165/) (2023)
- [Code World Models for General Game Playing](https://mlsanity.com/paper/2510.04542/) (2025)
- [Are ChatGPT and GPT-4 Good Poker Players? -- A Pre-Flop Analysis](https://mlsanity.com/paper/2308.12466/) (2023)
- [PokerGPT: An End-to-End Lightweight Solver for Multi-Player Texas Hold'em via Large Language Model](https://mlsanity.com/paper/2401.06781/) (2024)
- [PokerBench: Training Large Language Models to become Professional Poker Players](https://mlsanity.com/paper/2501.08328/) (2025)
- [Solving Large Sequential Games with the Excessive Gap Technique](https://mlsanity.com/paper/1810.03063/) (2018)
- [Pok'eChamp: an Expert-level Minimax Language Agent](https://mlsanity.com/paper/2503.04094/) (2025)
- [Measuring the Size of Large No-Limit Poker Games](https://mlsanity.com/paper/1302.7008/) (2013)
- [RL-CFR: Improving Action Abstraction for Imperfect Information Extensive-Form Games with Reinforcement Learning](https://mlsanity.com/paper/2403.04344/) (2024)
- [How Far Are LLMs from Professional Poker Players? Revisiting Game-Theoretic Reasoning with Agentic Tool Use](https://mlsanity.com/paper/2602.00528/) (2026)
- [Effective, Efficient, and General Information Abstraction for Imperfect-Information Extensive-Form Games](https://mlsanity.com/paper/2605.10900/) (2026)
- [From Natural Language to Extensive-Form Game Representations](https://mlsanity.com/paper/2501.17282/) (2025)
- [Real-Time Parallel Counterfactual Regret Minimization](https://mlsanity.com/paper/2605.19928/) (2026)
- [Correlated Chance Sampling for Monte Carlo Counterfactual Regret Minimization](https://mlsanity.com/paper/2607.27035/) (2026)
- [PokerSkill: LLMs Can Play Expert-Level Poker without Training or Solvers](https://mlsanity.com/paper/2605.30094/) (2026)
- [Can Large Language Models Master Complex Card Games?](https://mlsanity.com/paper/2509.01328/) (2025)
- [Agents That Certify Their Own Exploits: Confidence-Scheduled Restricted Responses for Safe Opponent Exploitation](https://mlsanity.com/paper/2607.28520/) (2026)
- [Solver-Guided Reasoning for Mixed-Equilibrium Strategies](https://mlsanity.com/paper/2608.06741/) (2026)
- [ICM Out! Better Tournament Strategy from Computed Continuations, vs. Solvers and LLMs](https://mlsanity.com/paper/2608.09586/) (2026)

## Built on by (most cited first, 1 of 1)

- [GPU-CFR: 80x Faster Counterfactual Regret Minimization by Compiling the Game to Static Dataflow and CUDA Graph Replay](https://mlsanity.com/paper/2609.11923/) (2026)

## Abstract (arXiv)

Information abstraction, which groups strategically similar private states into a tractable number of buckets, is essential for scaling game-solving algorithms to large imperfect-information games. Constructing effective abstractions, however, has traditionally required domain-specific evaluators such as hand-strength calculators or equity estimators, which demand expert knowledge and engineering effort and are unavailable for most less-studied games. We propose the Abstraction Agent, a zero-shot pipeline that uses a large language model (LLM) to discover continuous strategic features from a natural-language game description, score private states on these features, and cluster them into abstraction buckets, without any game-specific evaluator, training data, or game-tree traversal during abstraction construction. The pipeline runs in four phases: feature discovery with calibration anchors, batched private-state scoring, correlation-based feature selection, and k-means clustering. The resulting abstractions reduce lifted-strategy exploitability by up to 62% relative to an expected-hand-strength baseline on heads-up no-limit Texas hold'em (HUNL) turn endgames, and beat a scalar rank baseline at every granularity on ROVER Trials, an original game absent from any pretraining corpus. Beyond these quantitative benchmarks, the pipeline transfers with unchanged prompts to four-card Pot-Limit Omaha, HUNL preflop and flop, and Riichi Mahjong, where the discovered features track each game's recognized strategic concepts. This is structured knowledge elicitation: converting implicit strategic knowledge in LLM parameters into explicit numerical features for downstream algorithmic computation. The code is available at https://github.com/lbn187/AbstractionAgent.