Abstraction Agent

• URL: http://arxiv.org/abs/2609.04303v1
• Date: Thu, 03 Sep 2026 17:43:55 GMT
• Title: Abstraction Agent
• Abstract summary: Abstraction Agent builds effective information abstractions for imperfect information games.<n>The pipeline runs in four phases: feature discovery with calibration anchors, batched private-state scoring, correlation-based feature selection, and $k$-means clustering.<n>The pipeline reduces lifted-strategy exploitability by up 62% relative to an expected-hand-strength baseline.
• Score: 38.06569764716213
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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.

Related papers

• The Generalization Spectrum: A Chromatographic Approach to Evaluating Learning Algorithms [48.61395929803806]
  Traditional evaluations measure a learning algorithm's final performance on an i.i.d. test set, reducing learning to a single aggregate score.<n>This approach obscures a fundamental question: to what extent does learning from a specific example generalize to others?<n>We introduce the Generalization Spectrum, an evaluation framework designed to expose this hidden dimension.
  arXiv  Detail & Related papers  (2026-06-24T06:26:02Z)
• Domain-Independent Game Abstraction using Word Embedding Techniques [62.94836578759359]
  We propose a domain-independent approach to game abstraction.<n>Treating each action as a word and gameplay data as a corpus, word vectors can be trained to represent each action as a real-valued vector.
  arXiv  Detail & Related papers  (2026-05-15T02:30:36Z)
• KrwEmd: Revising the Imperfect-Recall Abstraction from Forgetting Everything [26.872387657059775]
  This paper presents KrwEmd, the first practical algorithm designed to address this problem.<n>We first introduce the k-recall winrate feature, which not only qualitatively distinguishes signal observation infosets by leveraging both future and, crucially, historical game information, but also quantitatively captures their similarity.<n>We then develop the KrwEmd algorithm, which clusters signal observation infosets using earth mover's distance to measure discrepancies between their features.
  arXiv  Detail & Related papers  (2025-11-15T08:15:50Z)
• Key-Augmented Neural Triggers for Knowledge Sharing [3.8718804267599416]
  Key-Augmented Neural Triggers embeds knowledge anchors into both training and inference.<n>It reduces fragmentation and grounding inference in localized context.<n>It is well-suited for scalable, low-latency, on-premise deployments.
  arXiv  Detail & Related papers  (2025-08-05T11:40:56Z)
• CountingDINO: A Training-free Pipeline for Class-Agnostic Counting using Unsupervised Backbones [7.717986156838291]
  Class-agnostic counting (CAC) aims to estimate the number of objects in images without being restricted to predefined categories.<n>Current exemplar-based CAC methods rely heavily on labeled data for training.<n>We introduce CountingDINO, the first training-free exemplar-based CAC framework.
  arXiv  Detail & Related papers  (2025-04-23T09:48:08Z)
• Strategy Game-Playing with Size-Constrained State Abstraction [44.99833362998488]
  Playing strategy games is a challenging problem for artificial intelligence (AI)<n>One of the major challenges is the large search space due to a diverse set of game components.<n>State abstraction has been applied to search-based game AI and has brought significant performance improvements.
  arXiv  Detail & Related papers  (2024-08-12T14:50:18Z)
• SPRING: Studying the Paper and Reasoning to Play Games [102.5587155284795]
  We propose a novel approach, SPRING, to read the game's original academic paper and use the knowledge learned to reason and play the game through a large language model (LLM) In experiments, we study the quality of in-context "reasoning" induced by different forms of prompts under the setting of the Crafter open-world environment. Our experiments suggest that LLMs, when prompted with consistent chain-of-thought, have great potential in completing sophisticated high-level trajectories.
  arXiv  Detail & Related papers  (2023-05-24T18:14:35Z)
• Active Learning for Abstractive Text Summarization [50.79416783266641]
  We propose the first effective query strategy for Active Learning in abstractive text summarization. We show that using our strategy in AL annotation helps to improve the model performance in terms of ROUGE and consistency scores.
  arXiv  Detail & Related papers  (2023-01-09T10:33:14Z)
• Elastic Monte Carlo Tree Search with State Abstraction for Strategy Game Playing [58.720142291102135]
  Strategy video games challenge AI agents with their search space caused by complex game elements. State abstraction is a popular technique that reduces the state space complexity. We propose Elastic MCTS, an algorithm that uses state abstraction to play strategy games.
  arXiv  Detail & Related papers  (2022-05-30T14:18:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.