语言大模型 / LLM

A Review of Developmental Interpretability in Large Language Models

Ihor Kendiukhov

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

Mechanistic Exploration of Backdoored Large Language Model Attention Patterns

Mohammed Abu Baker, Lakshmi Babu-Saheer

机构 * Department of Computer Science, Anglia Ruskin University（计算机科学系，安格利亚 Ruskin 大学） 纠错

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

Comments 13 pages. Mechanistic analysis of backdoored LLMs (Qwen2.5-3B). Code: https://github.com/mshahoyi/sa_attn_analysis. Base model: unsloth/Qwen2.5-3B-Instruct-unsloth-bnb-4bit. Finetuned models: https://huggingface.co/collections/mshahoyi/simple-sleeper-agents-68a1df3a7aaff310aa0e5336

Rotary Offset Features in Large Language Models

André Jonasson

机构 * André Jonasson（独立研究者） 纠错

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

Beyond Interpretability: Exploring the Comprehensibility of Adaptive Video Streaming through Large Language Models

Lianchen Jia, Chaoyang Li, Ziqi Yuan, Jiahui Chen, Tianchi Huang, Jiangchuan Liu, Lifeng Sun

机构 * Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系） ； School of Computing Science, Simon Fraser University（西蒙弗雷泽大学计算机科学系） ； Beijing National Research Center for Information Science and Technology（北京信息科学国家研究中心） 纠错

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

Comments ACM Multimedia2025

NEAT: Concept driven Neuron Attribution in LLMs

Vivek Hruday Kavuri, Gargi Shroff, Rahul Mishra

机构 * IIIT Hyderabad（IIIT海得拉巴） 纠错

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

Beyond Transcription: Mechanistic Interpretability in ASR

Neta Glazer, Yael Segal-Feldman, Hilit Segev, Aviv Shamsian, Asaf Buchnick, Gill Hetz, Ethan Fetaya, Joseph Keshet, Aviv Navon

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

Dancing with Deer: A Constructional Perspective on MWEs in the Era of LLMs

Claire Bonial, Julia Bonn, Harish Tayyar Madabushi

机构 * U.S. Army Research Lab（美国陆军研究实验室） ； University of Colorado Boulder（科罗拉多大学博尔德分校） ； University of Bath（巴斯大学） 纠错

专题命中 知识编辑与模型理解 ：large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Chapter in Phraseology and Multiword Expressions, Language Science Press (to appear)

UnPose: Uncertainty-Guided Diffusion Priors for Zero-Shot Pose Estimation

Zhaodong Jiang, Ashish Sinha, Tongtong Cao, Yuan Ren, Bingbing Liu, Binbin Xu

机构 * Huawei Noah’s Ark Lab, Canada（华为诺亚实验室，加拿大） ； University of Toronto, Canada（多伦多大学，加拿大） 纠错

专题命中 知识编辑与模型理解 ：foundation model(abstract)

Comments Published at the Conference on Robot Learning (CoRL) 2025. For more details please visit https://frankzhaodong.github.io/UnPose