본문

서브메뉴

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 Embodi...
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  
Large Language Models
키워드  
Reasoning
키워드  
Task and motion planning
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.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF16114 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.