서브메뉴
검색
Advancing Mathematical Reasoning With Language Models: A Multimodal and Knowledge-Intensive Perspective.
Advancing Mathematical Reasoning With Language Models: A Multimodal and Knowledge-Intensive Perspective.
상세정보
- 자료유형
- 학위논문
- Control Number
- 0017161940
- International Standard Book Number
- 9798382747545
- Dewey Decimal Classification Number
- 401
- Main Entry-Personal Name
- Lu, Pan.
- Publication, Distribution, etc. (Imprint
- [S.l.] : University of California, Los Angeles., 2024
- Publication, Distribution, etc. (Imprint
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- Physical Description
- 208 p.
- General Note
- Source: Dissertations Abstracts International, Volume: 85-11, Section: A.
- General Note
- Advisor: Chang, Kai-Wei.
- Dissertation Note
- Thesis (Ph.D.)--University of California, Los Angeles, 2024.
- Summary, Etc.
- 요약Mathematical reasoning is a pivotal component of human intelligence, crucial for advancing education and science. This dissertation delves into the development of language model systems capable of robust mathematical reasoning, marking a significant step toward realizing general artificial intelligence. We introduce multi-modal and knowledge-intensive benchmarks to assess the reasoning capabilities of large language models (LLMs) and vision-language models (VLMs) across real-world contexts, including visual information, tabular data, and scientific domains.This dissertation advances the field by proposing new pre-trained VLMs. For instance, Patch-Trm introduces a patch-based cross-modal Transformer model for abstract diagram reasoning. We also present innovative retrieval and tool-augmented algorithms that enhance LLM capabilities. Notably, Inter-GPS is a neuro-symbolic solver for geometry that demonstrates human-level performance, marking a first in the domain. Additionally, PromptPG pioneers the use of reinforcement learning for dynamic in-context example selection, significantly improving the stability and accuracy of LLMs. Another groundbreaking contribution is Chameleon, a model that integrates LLMs with external tools, vastly increasing their flexibility and effectiveness in real-world applications. The dissertation concludes by analyzing the latest advances in mathematical reasoning within visual contexts, and highlighting the current challenges and future prospects.
- Subject Added Entry-Topical Term
- Linguistics.
- Subject Added Entry-Topical Term
- Computer science.
- Index Term-Uncontrolled
- Language models
- Index Term-Uncontrolled
- Large language models
- Index Term-Uncontrolled
- Mathematical reasoning
- Index Term-Uncontrolled
- Vision language models
- Index Term-Uncontrolled
- Human intelligence
- Added Entry-Corporate Name
- University of California, Los Angeles Computer Science 0201
- Host Item Entry
- Dissertations Abstracts International. 85-11A.
- Electronic Location and Access
- 로그인을 한후 보실 수 있는 자료입니다.
- Control Number
- joongbu:657338
MARC
008250224s2024 us ||||||||||||||c||eng d■001000017161940
■00520250211151505
■006m o d
■007cr#unu||||||||
■020 ▼a9798382747545
■035 ▼a(MiAaPQ)AAI31298726
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a401
■1001 ▼aLu, Pan.
■24510▼aAdvancing Mathematical Reasoning With Language Models: A Multimodal and Knowledge-Intensive Perspective.
■260 ▼a[S.l.]▼bUniversity of California, Los Angeles. ▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a208 p.
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-11, Section: A.
■500 ▼aAdvisor: Chang, Kai-Wei.
■5021 ▼aThesis (Ph.D.)--University of California, Los Angeles, 2024.
■520 ▼aMathematical reasoning is a pivotal component of human intelligence, crucial for advancing education and science. This dissertation delves into the development of language model systems capable of robust mathematical reasoning, marking a significant step toward realizing general artificial intelligence. We introduce multi-modal and knowledge-intensive benchmarks to assess the reasoning capabilities of large language models (LLMs) and vision-language models (VLMs) across real-world contexts, including visual information, tabular data, and scientific domains.This dissertation advances the field by proposing new pre-trained VLMs. For instance, Patch-Trm introduces a patch-based cross-modal Transformer model for abstract diagram reasoning. We also present innovative retrieval and tool-augmented algorithms that enhance LLM capabilities. Notably, Inter-GPS is a neuro-symbolic solver for geometry that demonstrates human-level performance, marking a first in the domain. Additionally, PromptPG pioneers the use of reinforcement learning for dynamic in-context example selection, significantly improving the stability and accuracy of LLMs. Another groundbreaking contribution is Chameleon, a model that integrates LLMs with external tools, vastly increasing their flexibility and effectiveness in real-world applications. The dissertation concludes by analyzing the latest advances in mathematical reasoning within visual contexts, and highlighting the current challenges and future prospects.
■590 ▼aSchool code: 0031.
■650 4▼aLinguistics.
■650 4▼aComputer science.
■653 ▼aLanguage models
■653 ▼aLarge language models
■653 ▼aMathematical reasoning
■653 ▼aVision language models
■653 ▼aHuman intelligence
■690 ▼a0800
■690 ▼a0984
■690 ▼a0290
■71020▼aUniversity of California, Los Angeles▼bComputer Science 0201.
■7730 ▼tDissertations Abstracts International▼g85-11A.
■790 ▼a0031
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17161940▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
미리보기
내보내기
chatGPT토론
Ai 추천 관련 도서
ค้นหาข้อมูลรายละเอียด
- จองห้องพัก
- 캠퍼스간 도서대출
- 서가에 없는 책 신고
- โฟลเดอร์ของฉัน