Imagen de Portada

Enhancing Instruction Following of LLMs via Activation Steering with Dynamic Rejection

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Kang, Minjae, Kim, Jaehyung
Formato:	Preprint
Publicado:	2026
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Improving Instruction-Following in Language Models through Activation Steering
por: Stolfo, Alessandro, et al.
Publicado: (2024)

Few-shot Personalization of LLMs with Mis-aligned Responses
por: Kim, Jaehyung, et al.
Publicado: (2024)

Dynamically Scaled Activation Steering
por: Ferrando, Alex, et al.
Publicado: (2025)

Learning to Correct for QA Reasoning with Black-box LLMs
por: Kim, Jaehyung, et al.
Publicado: (2024)

Optimized Feature Generation for Tabular Data via LLMs with Decision Tree Reasoning
por: Nam, Jaehyun, et al.
Publicado: (2024)

Fast and Fluent Diffusion Language Models via Convolutional Decoding and Rejective Fine-tuning
por: Seo, Yeongbin, et al.
Publicado: (2025)

Debiasing Online Preference Learning via Preference Feature Preservation
por: Kim, Dongyoung, et al.
Publicado: (2025)

Training-free LLM Verification via Recycling Few-shot Examples
por: Lee, Dongseok, et al.
Publicado: (2025)

Global Evolutionary Steering: Refining Activation Steering Control via Cross-Layer Consistency
por: Jiang, Xinyan, et al.
Publicado: (2026)

CBMAS: Cognitive Behavioral Modeling via Activation Steering
por: Ismail, Ahmed H., et al.
Publicado: (2026)

Extracting Unlearned Information from LLMs with Activation Steering
por: Seyitoğlu, Atakan, et al.
Publicado: (2024)

Steering LLMs via Scalable Interactive Oversight
por: Zhou, Enyu, et al.
Publicado: (2026)

Self-Evolving LLMs via Continual Instruction Tuning
por: Kang, Jiazheng, et al.
Publicado: (2025)

Angular Steering: Behavior Control via Rotation in Activation Space
por: Vu, Hieu M., et al.
Publicado: (2025)

Is In-Context Learning Sufficient for Instruction Following in LLMs?
por: Zhao, Hao, et al.
Publicado: (2024)

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

Structural Reasoning Improves Molecular Understanding of LLM
por: Jang, Yunhui, et al.
Publicado: (2024)

Learning from the Undesirable: Robust Adaptation of Language Models without Forgetting
por: Nam, Yunhun, et al.
Publicado: (2025)

Minimizing Collateral Damage in Activation Steering
por: Nguyen, Tam, et al.
Publicado: (2026)

Steered LLM Activations are Non-Surjective
por: Mishra, Aayush, et al.
Publicado: (2026)

Activation Steering for Chain-of-Thought Compression
por: Azizi, Seyedarmin, et al.
Publicado: (2025)

Gap-K%: Measuring Top-1 Prediction Gap for Detecting Pretraining Data
por: Kwak, Minseo, et al.
Publicado: (2026)

Hierarchical Meta-Reinforcement Learning via Automated Macro-Action Discovery
por: Cho, Minjae, et al.
Publicado: (2024)

HyperSteer: Activation Steering at Scale with Hypernetworks
por: Sun, Jiuding, et al.
Publicado: (2025)

Enhancing and Assessing Instruction-Following with Fine-Grained Instruction Variants
por: Yang, Jiuding, et al.
Publicado: (2024)

Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMs
por: Han, Pengrui, et al.
Publicado: (2026)

The Road Less Traveled: Enhancing Exploration in LLMs via Sequential Sampling
por: Kang, Shijia, et al.
Publicado: (2025)

Steering Llama 2 via Contrastive Activation Addition
por: Panickssery, Nina, et al.
Publicado: (2023)

Steer Like the LLM: Activation Steering that Mimics Prompting
por: Heyman, Geert, et al.
Publicado: (2026)

Tabular Transfer Learning via Prompting LLMs
por: Nam, Jaehyun, et al.
Publicado: (2024)

Hierarchical Context Merging: Better Long Context Understanding for Pre-trained LLMs
por: Song, Woomin, et al.
Publicado: (2024)

Online Continual Learning For Interactive Instruction Following Agents
por: Kim, Byeonghwi, et al.
Publicado: (2024)

Personalized LLM Decoding via Contrasting Personal Preference
por: Bu, Hyungjune, et al.
Publicado: (2025)

The Rogue Scalpel: Activation Steering Compromises LLM Safety
por: Korznikov, Anton, et al.
Publicado: (2025)

Test-time Diverse Reasoning by Riemannian Activation Steering
por: Khanh, Ly Tran Ho, et al.
Publicado: (2025)

Steering Large Language Model Activations in Sparse Spaces
por: Bayat, Reza, et al.
Publicado: (2025)

Light-IF: Endowing LLMs with Generalizable Reasoning via Preview and Self-Checking for Complex Instruction Following
por: Wang, Chenyang, et al.
Publicado: (2025)

Steering Frozen LLMs: Adaptive Social Alignment via Online Prompt Routing
por: Zhang, Zeyu, et al.
Publicado: (2026)

Multi-property Steering of Large Language Models with Dynamic Activation Composition
por: Scalena, Daniel, et al.
Publicado: (2024)

Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering
por: Bigelow, Eric, et al.
Publicado: (2025)