Image de couverture de livre

One-shot Optimized Steering Vectors Mediate Safety-relevant Behaviors in LLMs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Dunefsky, Jacob, Cohan, Arman
Format:	Preprint
Publié:	2025
Sujets:	Machine Learning Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Observable Propagation: Uncovering Feature Vectors in Transformers
par: Dunefsky, Jacob, et autres
Publié: (2023)

On Evaluating LLM Alignment by Evaluating LLMs as Judges
par: Liu, Yixin, et autres
Publié: (2025)

SteeringSafety: A Systematic Safety Evaluation Framework of Representation Steering in LLMs
par: Siu, Vincent, et autres
Publié: (2025)

Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMs
par: Han, Pengrui, et autres
Publié: (2026)

Shifting Perspectives: Steering Vectors for Robust Bias Mitigation in LLMs
par: Siddique, Zara, et autres
Publié: (2025)

BarrierSteer: LLM Safety via Learning Barrier Steering
par: Tran, Thanh Q., et autres
Publié: (2026)

Dialz: A Python Toolkit for Steering Vectors
par: Siddique, Zara, et autres
Publié: (2025)

On the Non-Identifiability of Steering Vectors in Large Language Models
par: Venkatesh, Sohan, et autres
Publié: (2026)

COMAL: A Convergent Meta-Algorithm for Aligning LLMs with General Preferences
par: Liu, Yixin, et autres
Publié: (2024)

OPTIMA: Optimal One-shot Pruning for LLMs via Quadratic Programming Reconstruction
par: Mozaffari, Mohammad, et autres
Publié: (2025)

Understanding Reasoning in Thinking Language Models via Steering Vectors
par: Venhoff, Constantin, et autres
Publié: (2025)

Calibrating Long-form Generations from Large Language Models
par: Huang, Yukun, et autres
Publié: (2024)

The Rogue Scalpel: Activation Steering Compromises LLM Safety
par: Korznikov, Anton, et autres
Publié: (2025)

Graph-Regularized Sparse Autoencoders for LLM Safety Steering
par: Yeon, Jehyeok, et autres
Publié: (2025)

Enhancing LLM Steering through Sparse Autoencoder-Based Vector Refinement
par: Wang, Anyi, et autres
Publié: (2025)

Safety at One Shot: Patching Fine-Tuned LLMs with A Single Instance
par: Zhang, Jiawen, et autres
Publié: (2026)

Steering LLMs via Scalable Interactive Oversight
par: Zhou, Enyu, et autres
Publié: (2026)

One-Way Policy Optimization for Self-Evolving LLMs
par: Yang, Shuo, et autres
Publié: (2026)

One-shot Transfer Learning for Population Mapping
par: Shao, Erzhuo, et autres
Publié: (2021)

CBMAS: Cognitive Behavioral Modeling via Activation Steering
par: Ismail, Ahmed H., et autres
Publié: (2026)

Bridging Kolmogorov Complexity and Deep Learning: Asymptotically Optimal Description Length Objectives for Transformers
par: Shaw, Peter, et autres
Publié: (2025)

Angular Steering: Behavior Control via Rotation in Activation Space
par: Vu, Hieu M., et autres
Publié: (2025)

Examining Reasoning LLMs-as-Judges in Non-Verifiable LLM Post-Training
par: Liu, Yixin, et autres
Publié: (2026)

Improving Steering Vectors by Targeting Sparse Autoencoder Features
par: Chalnev, Sviatoslav, et autres
Publié: (2024)

Re-evaluating Automatic LLM System Ranking for Alignment with Human Preference
par: Gao, Mingqi, et autres
Publié: (2024)

One-shot Federated Learning Methods: A Practical Guide
par: Liu, Xiang, et autres
Publié: (2025)

Exploration Behavior of Untrained Policies
par: Adamczyk, Jacob
Publié: (2025)

Steering LLMs for Formal Theorem Proving
par: Kirtania, Shashank, et autres
Publié: (2025)

Enhancing Instruction Following of LLMs via Activation Steering with Dynamic Rejection
par: Kang, Minjae, et autres
Publié: (2026)

ExecTune: Effective Steering of Black-Box LLMs with Guide Models
par: Lingam, Vijay, et autres
Publié: (2026)

Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
par: Liu, Andy Zeyi, et autres
Publié: (2026)

References Improve LLM Alignment in Non-Verifiable Domains
par: Shi, Kejian, et autres
Publié: (2026)

Parametric Feature Transfer: One-shot Federated Learning with Foundation Models
par: Beitollahi, Mahdi, et autres
Publié: (2024)

ICPL: Few-shot In-context Preference Learning via LLMs
par: Yu, Chao, et autres
Publié: (2024)

COLD-Steer: Steering Large Language Models via In-Context One-step Learning Dynamics
par: Sharma, Kartik, et autres
Publié: (2026)

GuidedSampling: Steering LLMs Towards Diverse Candidate Solutions at Inference-Time
par: Handa, Divij, et autres
Publié: (2025)

Steering Frozen LLMs: Adaptive Social Alignment via Online Prompt Routing
par: Zhang, Zeyu, et autres
Publié: (2026)

Support Vector Boosting Machine (SVBM): Enhancing Classification Performance with AdaBoost and Residual Connections
par: Lian, Junbo Jacob
Publié: (2024)

Focus On This, Not That! Steering LLMs with Adaptive Feature Specification
par: Lamb, Tom A., et autres
Publié: (2024)

FedLPA: One-shot Federated Learning with Layer-Wise Posterior Aggregation
par: Liu, Xiang, et autres
Publié: (2023)