Skip to content
conftrace_

feature intervention

5 papers

Explore in graph

Co-occurring keywords · papers mentioning both

sparse autoencoder (2) gradient short-circuit (1) local gradient direction (1) inference-stage technique (1) large language model (1) prompt enrichment (1) model memorization (1) causal mediation (1) pii leakage (1) gradient analysis (1)

Papers

FreeCond: Free Lunch in the Input Conditions of Text-Guided Inpainting WACV 2026 Why Does Reinforcement Learning Generalize? A Feature-Level Mechanistic Study of Post-Training in Large Language Models ACL 2026 Gradient Short-Circuit: Efficient Out-of-Distribution Detection via Feature Intervention ICCV 2025 PrivacyScalpel: Enhancing LLM Privacy via Interpretable Feature Intervention with Sparse Autoencoders EMNLP 2025 Understanding Refusal in Language Models with Sparse Autoencoders EMNLP 2025