paper-with-me

Papers

Multi-property Steering of Large Language Models with Dynamic Activation Composition

2024-06-25 · Daniel Scalena, Gabriele Sarti, Malvina Nissim

Activation steering methods were shown to be effective in conditioning language model generation by additively intervening over models' intermediate representations. However, the evaluation of these techniques has so far been limited to single conditioning properties and synthetic settings. In this work, we conduct a comprehensive evaluation of various activation steering strategies, highlighting the property-dependent nature of optimal parameters to ensure a robust effect throughout generation. To address this issue, we propose Dynamic Activation Composition, an information-theoretic approach to modulate the steering intensity of one or more properties throughout generation. Our experiments on multi-property steering show that our method successfully maintains high conditioning while minimizing the impact of conditioning on generation fluency.

📄 PDF Abstract BibTeX arXiv:2406.17563

Code (1)

danielsc4/dynamic-activation-composition 공식 구현 pytorch

Tasks

Language ModelingLanguage Modelling

Similar Papers 제목 키워드 기반

DISCO: Disentangled Communication Steering for Large Language Models

2025-09-20 · Max Torop, Aria Masoomi, Masih Eskandar, Jennifer Dy arxiv

A variety of recent methods guide large language model outputs via the inference-time addition of steering vectors to residual-stream or attention-head representations. In contrast, we propose to inject steering vectors …

Steering Safely or Off a Cliff? Rethinking Specificity and Robustness in Inference-Time Interventions

2026-02-05 · Navita Goyal, Hal Daumé arxiv

Model steering, which involves intervening on hidden representations at inference time, has emerged as a lightweight alternative to finetuning for precisely controlling large language models. While steering efficacy has …

On the Non-Identifiability of Steering Vectors in Large Language Models

2026-02-06 · Sohan Venkatesh, Ashish Mahendran Kurapath arxiv

Activation steering methods are widely used to control large language model (LLM) behavior and are often interpreted as revealing meaningful internal representations. This interpretation assumes that steering directions …

Dynamic Multimodal Activation Steering for Hallucination Mitigation in Large Vision-Language Models

2026-02-25 · Jianghao Yin, Qin Chen, Kedi Chen, Jie Zhou 외 arxiv

Large Vision-Language Models (LVLMs) exhibit outstanding performance on vision-language tasks but struggle with hallucination problems. Through in-depth analysis of LVLM activation patterns, we reveal two key findings: 1…

Semantic Similarity

SLIM: Sparse Latent Steering for Interpretable and Property-Directed LLM-Based Molecular Editing

2026-05-11 · Mingxu Zhang, Yuhan Li, Lujundong Li, Dazhong Shen 외 arxiv

Large language models possess strong chemical reasoning capabilities, making them effective molecular editors. However, property-relevant information is implicitly entangled across their dense hidden states, providing no…