서브메뉴
검색
Enhancing the Reasoning Capabilities of Large Language Models: From Game Playing to Embodied Planning
Enhancing the Reasoning Capabilities of Large Language Models: From Game Playing to Embodied Planning
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105643
- ISBN
- 9798265483263
- DDC
- 310
- 저자명
- Wang, Shu.
- 서명/저자
- Enhancing the Reasoning Capabilities of Large Language Models: From Game Playing to Embodied Planning
- 발행사항
- [Sl] : University of California, Los Angeles, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 153 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-06, Section: B.
- 주기사항
- Advisor: Wu, Ying Nian.
- 학위논문주기
- Thesis (Ph.D.)--University of California, Los Angeles, 2025.
- 초록/해제
- 요약Reasoning is a fundamental capability of human intelligence that enables complex problem-solving and decision-making. Recent advances in Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks, yet their reasoning abilities in complex, long-horizon scenarios remain inadequately understood. This thesis systematically investigates and enhances the reasoning capabilities of LLMs across three progressively complex domains: constrained game playing, task and motion planning, and sequential manipulation in partially known environments.First, we explore LLM reasoning in the structured domain of chess, where we demonstrate that language explanations combining long-term strategic thinking and short-term tactical analysis significantly enhance model performance. Through the MATE dataset of 1 million annotated chess positions, we show that fine-tuned models incorporating strategy and tactic annotations outperform state-of-the-art commercial LLMs by 24.2%, establishing that explicit reasoning guidance improves decision-making quality.Second, we present LLM3 , a framework that leverages LLMs as domain-independent interfaces between symbolic task planning and continuous motion planning. By categorizing motion planning feedback into collision and unreachability failure modes, LLM3 enables iterative refinement of action sequences and parameters. Experimental results demonstrate a 91.3% success rate across realistic scenarios, with 36.1% reduction in travel distance compared to baselines, highlighting the effectiveness of LLM-based failure reasoning.Third, we introduce EPoG, which integrates exploration and sequential manipulation planning on graph-based scene representations for partially known environments. By employing LLMs for both informed exploration prioritization and situated replanning, EPoG naturally combines information gathering with task execution. Across 46 household scenes and 5 long-horizon tasks, EPoG achieves superior performance while reducing exploration overhead by 40% and travel distance by 36.2%.Collectively, this thesis makes three primary contributions: (1) demonstrating that language-based explanations enhance LLM reasoning across domains of increasing complexity, (2) establishing effective methods for integrating LLM reasoning with classical planning algorithms through structured feedback mechanisms, and (3) showing that LLMs can provide domain-independent heuristics that improve both planning efficiency and adaptability to environmental uncertainty. These findings advance our understanding of LLM reasoning capabilities and provide practical frameworks for deploying LLMs in complex embodied AI applications.
- 일반주제명
- Statistics
- 일반주제명
- Information technology
- 일반주제명
- Robotics
- 키워드
- Chess
- 키워드
- Game
- 키워드
- Reasoning
- 기타저자
- University of California, Los Angeles Statistics 0891
- 기본자료저록
- Dissertations Abstracts International. 87-06B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017360952
■00520260202105643
■006m o d
■007cr#unu||||||||
■020 ▼a9798265483263
■035 ▼a(MiAaPQ)AAI32399369
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a310
■1001 ▼aWang, Shu.
■24510▼aEnhancing the Reasoning Capabilities of Large Language Models: From Game Playing to Embodied Planning
■260 ▼a[Sl]▼bUniversity of California, Los Angeles▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a153 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-06, Section: B.
■500 ▼aAdvisor: Wu, Ying Nian.
■5021 ▼aThesis (Ph.D.)--University of California, Los Angeles, 2025.
■520 ▼aReasoning is a fundamental capability of human intelligence that enables complex problem-solving and decision-making. Recent advances in Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse tasks, yet their reasoning abilities in complex, long-horizon scenarios remain inadequately understood. This thesis systematically investigates and enhances the reasoning capabilities of LLMs across three progressively complex domains: constrained game playing, task and motion planning, and sequential manipulation in partially known environments.First, we explore LLM reasoning in the structured domain of chess, where we demonstrate that language explanations combining long-term strategic thinking and short-term tactical analysis significantly enhance model performance. Through the MATE dataset of 1 million annotated chess positions, we show that fine-tuned models incorporating strategy and tactic annotations outperform state-of-the-art commercial LLMs by 24.2%, establishing that explicit reasoning guidance improves decision-making quality.Second, we present LLM3 , a framework that leverages LLMs as domain-independent interfaces between symbolic task planning and continuous motion planning. By categorizing motion planning feedback into collision and unreachability failure modes, LLM3 enables iterative refinement of action sequences and parameters. Experimental results demonstrate a 91.3% success rate across realistic scenarios, with 36.1% reduction in travel distance compared to baselines, highlighting the effectiveness of LLM-based failure reasoning.Third, we introduce EPoG, which integrates exploration and sequential manipulation planning on graph-based scene representations for partially known environments. By employing LLMs for both informed exploration prioritization and situated replanning, EPoG naturally combines information gathering with task execution. Across 46 household scenes and 5 long-horizon tasks, EPoG achieves superior performance while reducing exploration overhead by 40% and travel distance by 36.2%.Collectively, this thesis makes three primary contributions: (1) demonstrating that language-based explanations enhance LLM reasoning across domains of increasing complexity, (2) establishing effective methods for integrating LLM reasoning with classical planning algorithms through structured feedback mechanisms, and (3) showing that LLMs can provide domain-independent heuristics that improve both planning efficiency and adaptability to environmental uncertainty. These findings advance our understanding of LLM reasoning capabilities and provide practical frameworks for deploying LLMs in complex embodied AI applications.
■590 ▼aSchool code: 0031.
■650 4▼aStatistics
■650 4▼aInformation technology
■650 4▼aRobotics
■653 ▼aChess
■653 ▼aGame
■653 ▼aLarge Language Models
■653 ▼aReasoning
■653 ▼aTask and motion planning
■690 ▼a0463
■690 ▼a0489
■690 ▼a0800
■690 ▼a0771
■71020▼aUniversity of California, Los Angeles▼bStatistics 0891.
■7730 ▼tDissertations Abstracts International▼g87-06B.
■790 ▼a0031
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360952▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


