No article text was recorded for this version.
activation patching
Version 1 of 1621 papers686 conceptsfrozen 2026-10-04
Papers in this version
- Blog: CMA-R: Causal Mediation Analysis for Explaining Rumour Detection
- Blog: Removing Spurious Correlation from Neural Network Interpretations
- Blog: How to use and interpret activation patching
- Blog: Causality Analysis for Evaluating the Security of Large Language Models
- Blog: Unveiling Language Competence Neurons: A Psycholinguistic Approach to Model Interpretability
- Blog: Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models
- Blog: CMA-R: Causal Mediation Analysis for Explaining Rumour Detection
- Blog: Causal Interpretation of Neural Network Computations with Contribution Decomposition
- Blog: Causal Mediation Analysis for Interpreting Large Language Models
- Blog: How Transformers Solve Propositional Logic Problems: A Mechanistic Analysis
- Blog: CausalGym: Benchmarking causal interpretability methods on linguistic tasks
- Blog: Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models
- Blog: RelP: Faithful and Efficient Circuit Discovery via Relevance Patching
- Blog: NeuroX: A Toolkit for Analyzing Individual Neurons in Neural Networks
- Blog: Dissecting Persona-Driven Reasoning in Language Models via Activation Patching
- Blog: Investigating Gender Bias in Language Models Using Causal Mediation Analysis
- Blog: SoK: a Comprehensive Causality Analysis Framework for Large Language Model Security
- Blog: Causality $\neq$ Invariance: Function and Concept Vectors in LLMs
- Blog: Granular Concept Circuits: Toward a Fine-Grained Circuit Discovery for Concept Representations
- Blog: Enhancing Multiple Dimensions of Trustworthiness in LLMs via Sparse Activation Control
- Blog: Induction Head Toxicity Mechanistically Explains Repetition Curse in Large Language Models
- Blog: MechIR: A Mechanistic Interpretability Framework for Information Retrieval
- Blog: An explainable transformer circuit for compositional generalization
- Blog: An Adversarial Example for Direct Logit Attribution: Memory Management in GELU-4L
- Blog: Images Speak Louder than Words: Understanding and Mitigating Bias in Vision-Language Model from a Causal Mediation Perspective
- Blog: Comparing Bottom-Up and Top-Down Steering Approaches on In-Context Learning Tasks
- Blog: Causal Learning for Socially Responsible AI
- Blog: Where’s the Plan? Locating Latent Planning in Language Models with Lightweight Mechanistic Interventions
- Blog: When Causal Intervention Meets Adversarial Examples and Image Masking for Deep Neural Networks
- Blog: Single-Position Intervention Fails: Distributed Output Templates Drive In-Context Learning
- Blog: Knowledge Circuits in Pretrained Transformers
- Blog: Emergent Symbol-like Number Variables in Artificial Neural Networks
- Blog: Causal Mediation Analysis for Interpreting Neural NLP: The Case of Gender Bias
- Blog: Identifying a Circuit for Verb Conjugation in GPT-2
- Blog: Surgical Activation Steering via Generative Causal Mediation
- Blog: Statistical Methods in Generative AI
- Blog: Understanding How Value Neurons Shape the Generation of Specified Values in LLMs
- Blog: Counterfactuals uncover the modular structure of deep generative models
- Blog: Increasing Trust in Language Models through the Reuse of Verified Circuits
- Blog: Causality-Based Neural Network Repair
- Blog: Towards Vision-Language Mechanistic Interpretability: A Causal Tracing Tool for BLIP
- Blog: Causal Analysis for Robust Interpretability of Neural Networks
- Blog: Generalized Wick Decompositions
- Blog: How Do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective
- Blog: Characterizing the Role of Similarity in the Property Inferences of Language Models
- Blog: Using Causal Analysis for Conceptual Deep Learning Explanation
- Blog: Language-Specific Representation of Emotion-Concept Knowledge Causally Supports Emotion Inference
- Blog: Causality-Based Neural Network Repair.
- Blog: Human Emotion Knowledge Representation Emerges in Large Language Model and Supports Discrete Emotion Inference
- Blog: Emergent Symbol-like Number Variables in Artificial Neural Networks
- Blog: Causal Attribution via Activation Patching
- Blog: How do Llamas process multilingual text? A latent exploration through activation patching
- Blog: When Models Decide and When They Bind: A Two-Stage Computation for Multiple-Choice Question-Answering
- Blog: The Hydra Effect: Emergent Self-repair in Language Model Computations
- Blog: How do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective
- Blog: How and where does CLIP process negation?
- Blog: What's in the Box? Exploring the Inner Life of Neural Networks with Robust Rules
- Blog: Induced Modularity and Community Detection for Functionally Interpretable Reinforcement Learning
- Blog: Does Knowledge Localization Hold True? Surprising Differences Between Entity and Relation Perspectives in Language Models
- Blog: How Do Multilingual Models Remember? Investigating Multilingual Factual Recall Mechanisms
Showing 60 of 621 papers.
History & citation
Current · Permanent link
- Mapping Neural Circuits via Causal Mediation merged into this direction (97% overlap)
- Mapping and Steering Causal Circuits in Language Models merged into this direction (96% overlap)
- Aligning Neural Networks with Symbolic Causal Models merged into this direction (50% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_activation_patching_f2ebb2db,
title = {activation patching},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/activation-patching/rdr_f2ebb2db14ec424b96173a2e4863ac48},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}This exact version (version 1, frozen text):
@misc{lacuna_activation_patching_f2ebb2db_v1,
title = {activation patching},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/activation-patching/rdr_f2ebb2db14ec424b96173a2e4863ac48/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}