Imagen de Portada

Why Steering Works: Toward a Unified View of Language Model Parameter Dynamics

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Xu, Ziwen, Wu, Chenyan, Sun, Hengyu, Hong, Haiwen, Wang, Mengru, Yao, Yunzhi, Huang, Longtao, Xue, Hui, Deng, Shumin, Chu, Zhixuan, Chen, Huajun, Zhang, Ningyu
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language Artificial Intelligence Computer Vision and Pattern Recognition Information Retrieval Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Knowledge Circuits in Pretrained Transformers
por: Yao, Yunzhi, et al.
Publicado: (2024)

Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms
por: Wang, Mengru, et al.
Publicado: (2025)

From Data to Behavior: Predicting Unintended Model Behaviors Before Training
por: Wang, Mengru, et al.
Publicado: (2026)

Towards A Unified View of Answer Calibration for Multi-Step Reasoning
por: Deng, Shumin, et al.
Publicado: (2023)

CaKE: Circuit-aware Editing Enables Generalizable Knowledge Learners
por: Yao, Yunzhi, et al.
Publicado: (2025)

Automating Steering for Safe Multimodal Large Language Models
por: Wu, Lyucheng, et al.
Publicado: (2025)

How Controllable Are Large Language Models? A Unified Evaluation across Behavioral Granularities
por: Xu, Ziwen, et al.
Publicado: (2026)

LLMs for Knowledge Graph Construction and Reasoning: Recent Capabilities and Future Opportunities
por: Zhu, Yuqi, et al.
Publicado: (2023)

Editing Conceptual Knowledge for Large Language Models
por: Wang, Xiaohan, et al.
Publicado: (2024)

WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models
por: Wang, Peng, et al.
Publicado: (2024)

EasyEdit2: An Easy-to-use Steering Framework for Editing Large Language Models
por: Xu, Ziwen, et al.
Publicado: (2025)

EasyEdit: An Easy-to-use Knowledge Editing Framework for Large Language Models
por: Wang, Peng, et al.
Publicado: (2023)

Exploring Model Kinship for Merging Large Language Models
por: Hu, Yedi, et al.
Publicado: (2024)

CKnowEdit: A New Chinese Knowledge Editing Dataset for Linguistics, Facts, and Logic Error Correction in LLMs
por: Fang, Jizhan, et al.
Publicado: (2024)

Detoxifying Large Language Models via Knowledge Editing
por: Wang, Mengru, et al.
Publicado: (2024)

ReCode: Updating Code API Knowledge with Reinforcement Learning
por: Wu, Haoze, et al.
Publicado: (2025)

Two Experts Are All You Need for Steering Thinking: Reinforcing Cognitive Effort in MoE Reasoning Models Without Additional Training
por: Wang, Mengru, et al.
Publicado: (2025)

OneEdit: A Neural-Symbolic Collaboratively Knowledge Editing System
por: Zhang, Ningyu, et al.
Publicado: (2024)

Information Extraction in Low-Resource Scenarios: Survey and Perspective
por: Deng, Shumin, et al.
Publicado: (2022)

How LoRA Remembers? A Parametric Memory Law for LLM Finetuning
por: Xu, Ziwen, et al.
Publicado: (2026)

LightMem: Lightweight and Efficient Memory-Augmented Generation
por: Fang, Jizhan, et al.
Publicado: (2025)

Unveiling the Pitfalls of Knowledge Editing for Large Language Models
por: Li, Zhoubo, et al.
Publicado: (2023)

How Do LLMs Acquire New Knowledge? A Knowledge Circuits Perspective on Continual Pre-Training
por: Ou, Yixin, et al.
Publicado: (2025)

Unified Hallucination Detection for Multimodal Large Language Models
por: Chen, Xiang, et al.
Publicado: (2024)

QExplorer: Large Language Model Based Query Extraction for Toxic Content Exploration
por: Ren, Shaola, et al.
Publicado: (2025)

Retrieval-augmented Prompt Learning for Pre-trained Foundation Models
por: Chen, Xiang, et al.
Publicado: (2025)

EasyInstruct: An Easy-to-use Instruction Processing Framework for Large Language Models
por: Ou, Yixin, et al.
Publicado: (2024)

SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research
por: Qiao, Shuofei, et al.
Publicado: (2026)

Knowledge Mechanisms in Large Language Models: A Survey and Perspective
por: Wang, Mengru, et al.
Publicado: (2024)

OneGen: Efficient One-Pass Unified Generation and Retrieval for LLMs
por: Zhang, Jintian, et al.
Publicado: (2024)

StructMem: Structured Memory for Long-Horizon Behavior in LLMs
por: Xu, Buqiang, et al.
Publicado: (2026)

Making Language Models Better Tool Learners with Execution Feedback
por: Qiao, Shuofei, et al.
Publicado: (2023)

ResearchArena: Benchmarking Large Language Models' Ability to Collect and Organize Information as Research Agents
por: Kang, Hao, et al.
Publicado: (2024)

Why Do Open-Source LLMs Struggle with Data Analysis? A Systematic Empirical Study
por: Zhu, Yuqi, et al.
Publicado: (2025)

Dwell in the Beginning: How Language Models Embed Long Documents for Dense Retrieval
por: Coelho, João, et al.
Publicado: (2024)

IEPile: Unearthing Large-Scale Schema-Based Information Extraction Corpus
por: Gui, Honghao, et al.
Publicado: (2024)

RaFe: Ranking Feedback Improves Query Rewriting for RAG
por: Mao, Shengyu, et al.
Publicado: (2024)

Towards Human-centered Proactive Conversational Agents
por: Deng, Yang, et al.
Publicado: (2024)

Modality and Task Adaptation for Enhanced Zero-shot Composed Image Retrieval
por: Li, Haiwen, et al.
Publicado: (2024)

CodeKGC: Code Language Model for Generative Knowledge Graph Construction
por: Bi, Zhen, et al.
Publicado: (2023)