Cover Image

Denoising Concept Vectors with Sparse Autoencoders for Improved Language Model Steering

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhao, Haiyan, Wu, Xuansheng, Yang, Fan, Shen, Bo, Liu, Ninghao, Du, Mengnan
Format:	Preprint
Published:	2025
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Beyond Input Activations: Identifying Influential Latents by Gradient Sparse Autoencoders
by: Shu, Dong, et al.
Published: (2025)

A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models
by: Shu, Dong, et al.
Published: (2025)

Beyond Single Concept Vector: Modeling Concept Subspace in LLMs with Gaussian Distribution
by: Zhao, Haiyan, et al.
Published: (2024)

SAIF: A Sparse Autoencoder Framework for Interpreting and Steering Instruction Following of Language Models
by: He, Zirui, et al.
Published: (2025)

Enhancing LLM Steering through Sparse Autoencoder-Based Vector Refinement
by: Wang, Anyi, et al.
Published: (2025)

Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders
by: Wu, Xuansheng, et al.
Published: (2025)

Improving Steering Vectors by Targeting Sparse Autoencoder Features
by: Chalnev, Sviatoslav, et al.
Published: (2024)

Applying Large Language Models and Chain-of-Thought for Automatic Scoring
by: Lee, Gyeong-Geon, et al.
Published: (2023)

AutoSCORE: Enhancing Automated Scoring with Multi-Agent Large Language Models via Structured Component Recognition
by: Wang, Yun, et al.
Published: (2025)

RepIt: Steering Language Models with Concept-Specific Refusal Vectors
by: Siu, Vincent, et al.
Published: (2025)

BRIDGE the Gap: Mitigating Bias Amplification in Automated Scoring of English Language Learners via Inter-group Data Augmentation
by: Wang, Yun, et al.
Published: (2026)

Self-Regularization with Sparse Autoencoders for Controllable LLM-based Classification
by: Wu, Xuansheng, et al.
Published: (2025)

Retrieval-enhanced Knowledge Editing in Language Models for Multi-Hop Question Answering
by: Shi, Yucheng, et al.
Published: (2024)

Exploring Multilingual Probing in Large Language Models: A Cross-Language Analysis
by: Li, Daoyang, et al.
Published: (2024)

Language Ranker: A Metric for Quantifying LLM Performance Across High and Low-Resource Languages
by: Li, Zihao, et al.
Published: (2024)

Improving LLM Reasoning through Interpretable Role-Playing Steering
by: Wang, Anyi, et al.
Published: (2025)

From Language Modeling to Instruction Following: Understanding the Behavior Shift in LLMs after Instruction Tuning
by: Wu, Xuansheng, et al.
Published: (2023)

Could Small Language Models Serve as Recommenders? Towards Data-centric Cold-start Recommendations
by: Wu, Xuansheng, et al.
Published: (2023)

Exploring Concept Depth: How Large Language Models Acquire Knowledge and Concept at Different Layers?
by: Jin, Mingyu, et al.
Published: (2024)

Controllable LLM Reasoning via Sparse Autoencoder-Based Steering
by: Fang, Yi, et al.
Published: (2026)

The Impact of Steering Large Language Models with Persona Vectors in Educational Applications
by: Wu, Yongchao, et al.
Published: (2026)

The Impact of Reasoning Step Length on Large Language Models
by: Jin, Mingyu, et al.
Published: (2024)

Sparse Shift Autoencoders for Identifying Concepts from Large Language Model Activations
by: Joshi, Shruti, et al.
Published: (2025)

Concept-Centric Token Interpretation for Vector-Quantized Generative Models
by: Yang, Tianze, et al.
Published: (2025)

Artificial Intelligence Bias on English Language Learners in Automatic Scoring
by: Guo, Shuchen, et al.
Published: (2025)

Rep2Text: Decoding Full Text from a Single LLM Token Representation
by: Zhao, Haiyan, et al.
Published: (2025)

Towards Uncovering How Large Language Model Works: An Explainability Perspective
by: Zhao, Haiyan, et al.
Published: (2024)

SAEMark: Steering Personalized Multilingual LLM Watermarks with Sparse Autoencoders
by: Yu, Zhuohao, et al.
Published: (2025)

Fine-Grained Interpretation of Political Opinions in Large Language Models
by: Hu, Jingyu, et al.
Published: (2025)

Mechanistic Knobs in LLMs: Retrieving and Steering High-Order Semantic Features via Sparse Autoencoders
by: Zhang, Ruikang, et al.
Published: (2026)

Steering LLMs? Actually, Sparse Autoencoders can outperform simple baselines
by: Jørgensen, Mikkel Godsk, et al.
Published: (2026)

SparseRM: A Lightweight Preference Modeling with Sparse Autoencoder
by: Liu, Dengcan, et al.
Published: (2025)

Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders
by: Li, Aaron J., et al.
Published: (2025)

Use Sparse Autoencoders to Discover Unknown Concepts, Not to Act on Known Concepts
by: Peng, Kenny, et al.
Published: (2025)

Steering LVLMs via Sparse Autoencoder for Hallucination Mitigation
by: Hua, Zhenglin, et al.
Published: (2025)

A Survey on Fairness in Large Language Models
by: Li, Yingji, et al.
Published: (2023)

ASRU: Activation Steering Meets Reinforcement Unlearning for Multimodal Large Language Models
by: Guang, Jiahui, et al.
Published: (2026)

Causal Language Control in Multilingual Transformers via Sparse Feature Steering
by: Chou, Cheng-Ting, et al.
Published: (2025)

SAE-SSV: Supervised Steering in Sparse Representation Spaces for Reliable Control of Language Models
by: He, Zirui, et al.
Published: (2025)

BILLY: Steering Large Language Models via Merging Persona Vectors for Creative Generation
by: Pai, Tsung-Min, et al.
Published: (2025)