语言大模型 / LLM - arXivDaily 专题

2404.08636 2024-04-15 cs.CV 82%

Probing the 3D Awareness of Visual Foundation Models

Mohamed El Banani, Amit Raj, Kevis-Kokitsi Maninis, Abhishek Kar, Yuanzhen Li, Michael Rubinstein, Deqing Sun, Leonidas Guibas, Justin Johnson, Varun Jampani

专题命中知识编辑与模型理解：foundation model(title,abstract)；pretraining(abstract)

Comments Accepted to CVPR 2024. Project page: https://github.com/mbanani/probe3d

2403.19837 2024-04-12 cs.LG cs.AI cs.CL cs.CV cs.LO 82%

Concept-based Analysis of Neural Networks via Vision-Language Models

Ravi Mangal, Nina Narodytska, Divya Gopinath, Boyue Caroline Hu, Anirban Roy, Susmit Jha, Corina Pasareanu

专题命中知识编辑与模型理解：language model(title)；foundation model(abstract)；分类 cs.CL、cs.AI、cs.LG

2402.12865 2024-02-21 cs.CL cs.AI cs.LG 82%

Backward Lens: Projecting Language Model Gradients into the Vocabulary Space

Shahar Katz, Yonatan Belinkov, Mor Geva, Lior Wolf

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

2402.10978 2024-02-20 cs.LG cs.AI cs.CL 82%

Language Models with Conformal Factuality Guarantees

Christopher Mohri, Tatsunori Hashimoto

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

2309.16042 2024-01-18 cs.LG cs.AI cs.CL 82%

Towards Best Practices of Activation Patching in Language Models: Metrics and Methods

Fred Zhang, Neel Nanda

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 27 pages. ICLR 2024

2311.17618 2023-12-04 cs.CV 82%

ShapeGPT: 3D Shape Generation with A Unified Multi-modal Language Model

Fukun Yin, Xin Chen, Chi Zhang, Biao Jiang, Zibo Zhao, Jiayuan Fan, Gang Yu, Taihao Li, Tao Chen

专题命中知识编辑与模型理解：language model(title,abstract)；large language model(abstract)

2311.05729 2023-11-13 cs.CV 82%

GIPCOL: Graph-Injected Soft Prompting for Compositional Zero-Shot Learning

Guangyue Xu, Joyce Chai, Parisa Kordjamshidi

专题命中知识编辑与模型理解：prompting(title,abstract)；language model(abstract)

Comments WACV24

2305.00586 2023-11-03 cs.CL cs.AI cs.LG 82%

How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model

Michael Hanna, Ollie Liu, Alexandre Variengien

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments NeurIPS 2023

2310.14993 2023-10-24 cs.LG cs.AI cs.CL 82%

Understanding the Inner Workings of Language Models Through Representation Dissimilarity

Davis Brown, Charles Godfrey, Nicholas Konz, Jonathan Tu, Henry Kvinge

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments EMNLP 2023 (main)

2306.05077 2023-06-09 cs.CL cs.AI cs.LG 82%

Improving Language Model Integration for Neural Machine Translation

Christian Herold, Yingbo Gao, Mohammad Zeineldeen, Hermann Ney

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments accepted at ACL2023 (Findings)

2305.13073 2023-05-30 cs.CL cs.AI cs.DB cs.LG 82%

Text-to-SQL Error Correction with Language Models of Code

Ziru Chen, Shijie Chen, Michael White, Raymond Mooney, Ali Payani, Jayanth Srinivasa, Yu Su, Huan Sun

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments ACL 2023 Short Paper

2206.04624 2023-03-03 cs.CL cs.AI cs.CY cs.LG 82%

Factuality Enhanced Language Models for Open-Ended Text Generation

Nayeon Lee, Wei Ping, Peng Xu, Mostofa Patwary, Pascale Fung, Mohammad Shoeybi, Bryan Catanzaro

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments NeurIPS 2022

2203.16634 2022-12-07 cs.CL cs.AI cs.LG 82%

Transformer Language Models without Positional Encodings Still Learn Positional Information

Adi Haviv, Ori Ram, Ofir Press, Peter Izsak, Omer Levy

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Findings of EMNLP 2022

2206.11719 2022-09-13 cs.CL cs.AI cs.LG cs.PL cs.SE 82%

AST-Probe: Recovering abstract syntax trees from hidden representations of pre-trained language models

José Antonio Hernández López, Martin Weyssow, Jesús Sánchez Cuadrado, Houari Sahraoui

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

2108.11193 2022-06-09 cs.CL cs.AI cs.LG 82%

Models In a Spelling Bee: Language Models Implicitly Learn the Character Composition of Tokens

Itay Itzhak, Omer Levy

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments NAACL 2022

2202.10419 2022-05-24 cs.CL cs.AI cs.LG 82%

Interpreting Language Models with Contrastive Explanations

Kayo Yin, Graham Neubig

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

2110.02058 2022-03-14 cs.CL cs.AI cs.LG 82%

Interactively Providing Explanations for Transformer Language Models

Felix Friedrich, Patrick Schramowski, Christopher Tauchmann, Kristian Kersting

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

2104.05837 2021-09-13 cs.CL cs.AI cs.LG 82%

Relational World Knowledge Representation in Contextual Language Models: A Review

Tara Safavi, Danai Koutra

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments EMNLP 2021

2106.01950 2021-06-04 cs.CL cs.AI cs.LG 82%

The Case for Translation-Invariant Self-Attention in Transformer-Based Language Models

Ulme Wennberg, Gustav Eje Henter

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 11 pages, 8 figures, Accepted to ACL 2021

1909.04625 2019-09-11 cs.CL cs.AI cs.LG 82%

Representation of Constituents in Neural Language Models: Coordination Phrase as a Case Study

Aixiu An, Peng Qian, Ethan Wilcox, Roger Levy

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Comments To appear at EMNLP 2019

2606.11657 2026-06-11 cs.LG cs.AI 新提交 82%

Sparse probes and murky physics: a case study of interpretability challenges in a foundation model for continuum dynamics

稀疏探针与模糊物理：连续介质动力学基础模型可解释性挑战的案例研究

Katherine Rosenfeld, Maike Sonnewald

机构 * Gates Foundation（盖茨基金会）； UC Davis（加州大学戴维斯分校）

专题命中知识编辑与模型理解：foundation model(title,abstract)；分类 cs.AI、cs.LG

AI总结本研究通过稀疏自编码器探针分析连续介质动力学基础模型Walrus的内部机制，发现其内部特征与物理分解不完全一致，并存在输出级偏差，揭示了科学基础模型可解释性的关键挑战。

Comments 8 pages, 5 figures

Journal ref ICLR 2026 Workshop on Foundation Models for Science

2211.16327 2026-04-27 cs.AI cs.LG 82%

On the Power of Foundation Models

基础模型的威力

Yang Yuan

机构 * IIIS, Tsinghua University（清华大学信息学院）； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）； Shanghai Qi Zhi Institute（上海启智研究院）

专题命中知识编辑与模型理解：foundation model(title,abstract)；分类 cs.AI、cs.LG；LLM(comments)

AI总结本文通过范畴论探讨基础模型在提示学习和微调中的能力限制及泛化理论，提出新的泛化定理。

Comments ICML'23. This version polished paper with the help of LLM, fixed a few notational issues

2510.08931 2025-10-13 cs.AI cs.LG 82%

RADAR: Mechanistic Pathways for Detecting Data Contamination in LLM Evaluation

Ashish Kattamuri, Harshwardhan Fartale, Arpita Vats, Rahul Raja, Ishita Prasad

机构 * Proofpoint ； Indian Institute of Science（印度科学研究院）； Linkedin ； Meta FAIR

专题命中知识编辑与模型理解：LLM(title,abstract)；分类 cs.AI、cs.LG

Comments NeurIPS 2025 Workshop on Evaluating the Evolving LLM Lifecycle: Benchmarks, Emergent Abilities, and Scaling

2509.05801 2025-10-07 cs.LG cs.AI 82%

time2time: Causal Intervention in Hidden States to Simulate Rare Events in Time Series Foundation Models

Debdeep Sanyal, Aaryan Nagpal, Dhruv Kumar, Murari Mandal, Saurabh Deshpande

机构 * Birla AI Labs, Office of Ananya Birla（Birla AI实验室，Ananya Birla办公室）； BITS Pilani ； KIIT Bhubaneswar（KIIT巴尔班格斯）

专题命中知识编辑与模型理解：foundation model(title,abstract)；分类 cs.AI、cs.LG

Journal ref NeurIPS 2025 Workshop on Recent Advances in Time Series Foundation Models (BERT2S)

2410.13779 2025-05-21 cs.CL cs.LG 82%

The Mystery of the Pathological Path-star Task for Language Models

Arvid Frydenlund

机构 * University of Toronto（多伦多大学）； Vector Institute（向量研究所）

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.LG

Comments EMNLP 2024 Main at https://aclanthology.org/2024.emnlp-main.695/ See 'Language Models, Graph Searching, and Supervision Adulteration: When More Supervision is Less and How to Make More More' for a follow-up work

2502.18499 2025-02-27 cs.SE cs.AI cs.CL 82%

Mechanistic Understanding of Language Models in Syntactic Code Completion

Samuel Miller, Daking Rai, Ziyu Yao

专题命中知识编辑与模型理解：language model(title,abstract)；分类 cs.CL、cs.AI；foundation model(comments)

Comments 10 pages, 4 figures, accepted to the AAAI 2025 Workshop on Towards Knowledgeable Foundation Models

2412.14097 2024-12-19 cs.LG cs.AI cs.CV 82%

Adaptive Concept Bottleneck for Foundation Models Under Distribution Shifts

Jihye Choi, Jayaram Raghuram, Yixuan Li, Somesh Jha

专题命中知识编辑与模型理解：foundation model(title,abstract)；分类 cs.AI、cs.LG

Comments The preliminary version of the work appeared in the ICML 2024 Workshop on Foundation Models in the Wild

2407.08065 2024-07-12 cs.RO cs.AI cs.LG 82%

Towards Interpretable Foundation Models of Robot Behavior: A Task Specific Policy Generation Approach

Isaac Sheidlower, Reuben Aronson, Elaine Schaertl Short

专题命中知识编辑与模型理解：foundation model(title,abstract)；分类 cs.AI、cs.LG

Comments Short Paper accepted to RLC 2024 Workshop on Training Agents with Foundation Models

2608.25382 2026-08-27 cs.HC cs.AI cs.IR 新提交 81%

Q&A or Document-Based? The Effects of Interface Type on How Screen Reader Users Access Interconnected Documents

问答式还是基于文档式？界面类型对屏幕阅读器用户访问互联文档的影响

Colleen F. Cipriano, Yichun Zhao, Miguel A. Nacenta, Kotaro Hara, Jaylee Soh

专题命中知识编辑与模型理解：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)；分类 cs.AI

AI总结该研究对比问答界面（QAI）与文档界面（DI）对16名屏幕阅读器盲/低视力用户探索虚构世界的影响，发现DI更利于知识构建，但多数用户仍偏好QAI且存在认知偏差，凸显QAI访问信息空间的风险。

Comments 17 pages, 12 figures, accepted at ASSETS 2026

2608.22356 2026-08-25 cs.AI cs.HC 新提交 81%

Addressing the Selection Problem in Explainable AI

解决可解释人工智能中的选择问题

Claire Vlases, Katelyn Morrison

机构 * Carnegie Mellon University（卡内基梅隆大学）

专题命中知识编辑与模型理解：LLM(summary_cn,abstract)；分类 cs.AI

AI总结针对可解释人工智能（XAI）中用户难以选择合适技术的问题，研究提出多智能体大语言模型（LLM）编排工具，将用户查询转化为对应XAI解释技术，以解决选择问题。

Comments Accepted to the Workshop on Explainable Artificial Intelligence at the International Joint Conference on Artificial Intelligence 2026 (XAI@IJCAI26)

1. 知识编辑与模型理解 7608 篇

Probing the 3D Awareness of Visual Foundation Models

Concept-based Analysis of Neural Networks via Vision-Language Models

Backward Lens: Projecting Language Model Gradients into the Vocabulary Space

Language Models with Conformal Factuality Guarantees

Towards Best Practices of Activation Patching in Language Models: Metrics and Methods

ShapeGPT: 3D Shape Generation with A Unified Multi-modal Language Model

GIPCOL: Graph-Injected Soft Prompting for Compositional Zero-Shot Learning

How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model

Understanding the Inner Workings of Language Models Through Representation Dissimilarity

Improving Language Model Integration for Neural Machine Translation

Text-to-SQL Error Correction with Language Models of Code

Factuality Enhanced Language Models for Open-Ended Text Generation

Transformer Language Models without Positional Encodings Still Learn Positional Information

AST-Probe: Recovering abstract syntax trees from hidden representations of pre-trained language models

Models In a Spelling Bee: Language Models Implicitly Learn the Character Composition of Tokens

Interpreting Language Models with Contrastive Explanations

Interactively Providing Explanations for Transformer Language Models

Relational World Knowledge Representation in Contextual Language Models: A Review

The Case for Translation-Invariant Self-Attention in Transformer-Based Language Models

Representation of Constituents in Neural Language Models: Coordination Phrase as a Case Study

Sparse probes and murky physics: a case study of interpretability challenges in a foundation model for continuum dynamics

On the Power of Foundation Models

RADAR: Mechanistic Pathways for Detecting Data Contamination in LLM Evaluation

time2time: Causal Intervention in Hidden States to Simulate Rare Events in Time Series Foundation Models

The Mystery of the Pathological Path-star Task for Language Models

Mechanistic Understanding of Language Models in Syntactic Code Completion

Adaptive Concept Bottleneck for Foundation Models Under Distribution Shifts

Towards Interpretable Foundation Models of Robot Behavior: A Task Specific Policy Generation Approach

Q&A or Document-Based? The Effects of Interface Type on How Screen Reader Users Access Interconnected Documents

Addressing the Selection Problem in Explainable AI