When Attribution Patching Lies: Diagnosis and a Second-Order Correction
cs.LG, cs.AI
Submitted: 2026-06-05
Updated: 2026-11-05
Code: https://github.com/TransformerLensOrg/TransformerLens
Terminology
Sources
- AtP*: An efficient and scalable method for localizing LLM behaviour to components
- RelP: Faithful and Efficient Circuit Discovery in Language Models via Relevance Patching
- How to use and interpret activation patching
- Towards Best Practices of Activation Patching in Language Models: Metrics and Methods
- Emotion Concepts and their Function in a Large Language Model
- Gemma 2: Improving Open Language Models at a Practical Size
- The Llama 3 Herd of Models
Related papers
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis
- (Mis)generalization of Helpful-Only Fine-tuning