Whose Effect Is It? How Objectives Inflate Learned Subspace Interventions
cs.CL
Submitted: 2026-11-03
Updated: 2026-11-03
Code: https://github.com/tanyelai/whose-effect
Terminology
Sources
- Gemma 2: Improving Open Language Models at a Practical Size
- Localizing Model Behavior with Path Patching
- The Llama 3 Herd of Models
- How to use and interpret activation patching
- Routing Subspaces: Auditing Evaluation-to-Deployment Mismatch in Fine-Tuned Language Models
- AtP*: An efficient and scalable method for localizing LLM behaviour to components
- Qwen2.5 Technical Report
- A Reply to Makelov et al. (2023)'s "Interpretability Illusion" Arguments
Related papers
- ALARM: Audio-Language Alignment for Reasoning Models
- ThinkSum: Probabilistic reasoning over sets using large language models
- Coherence boosting: When your pretrained language model is not paying enough attention
- From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents
- Improving General Role-Playing Agents via Psychology-Grounded Reasoning and Role-Aware Policy Optimization
- The Art of Asking: Multilingual Prompt Optimization for Synthetic Data