WriteSAE introduces sparse autoencoders with rank-1 matrix atoms for recurrent state updates, allowing replacement tests that outperform deletion on 92.4% of positions and a formula predicting logit changes with R²=0.98.
hub Mixed citations
Localizing Model Behavior with Path Patching
Mixed citation behavior. Most common role is background (50%).
24
Pith papers citing it
Background
50% of classified citations
abstract
Localizing behaviors of neural networks to a subset of the network's components or a subset of interactions between components is a natural first step towards analyzing network mechanisms and possible failure modes. Existing work is often qualitative and ad-hoc, and there is no consensus on the appropriate way to evaluate localization claims. We introduce path patching, a technique for expressing and quantitatively testing a natural class of hypotheses expressing that behaviors are localized to a set of paths. We refine an explanation of induction heads, characterize a behavior of GPT-2, and open source a framework for efficiently running similar experiments.
hub tools
citation-role summary
background 3
method 3
citation-polarity summary
representative citing papers
Introduces Causal Functional Signatures grounded in causal evidence and ILP-learned architectural signatures to enable explicit, comparable, and portable mechanistic claims across model scales.
LLMs encode repeated token counts correctly in residual streams but a format-triggered MLP at 88-93% depth overwrites it with an incorrect fixed value.
In-context learning binds model outputs to the demonstrated label tokens as an exhaustive vocabulary, overriding semantic plausibility and causing fixation even with homogeneous or nonsense labels.
Transformer circuits show free evolution during SFT, rendering static mechanistic localization inadequate for future parameter updates due to inherent temporal latency.
The Linear Centroids Hypothesis reframes network features as directions in centroid spaces of local affine experts, unifying interpretability methods and yielding sparser, more faithful dictionaries, circuits, and saliency maps.
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
Vision-language models contain identifiable grounding and hallucination pathways; suppressing the latter reduces object hallucinations by up to 76% while preserving accuracy.
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
CoT traces align with internal answer commitment in only 61.9% of steps on average, dominated by confabulated continuations after commitment has stabilized.
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
Patch-effect graphs built from causal mediation, partial correlation, and co-influence, when analyzed with graph kernels, preserve task-discriminative signals from activation patching that outperform global shape descriptors and raw baselines on GPT-2 Small.
Token-level contrastive attribution yields informative signals for some LLM benchmark failures but is not universally applicable across datasets and models.
Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.
AP-MAE reconstructs masked attention patterns in LLMs with high accuracy, generalizes across models, predicts generation correctness at 55-70%, and enables 13.6% accuracy gains via targeted interventions.
An empirical audit identifies a strong SAE feature correlate for GPT-2 small failures on 'keys' prompts in the IOI task, performs ablation and baseline controls showing it is not causal, and presents the audit pipeline as the primary contribution.
Overthinking in medical QA is linearly decodable at 71.6% accuracy yet fixed residual-stream steering yields no correction across 29 configurations, while enabling selective abstention with AUROC 0.610.
The survey organizes mechanistic interpretability techniques into a Locate-Steer-Improve framework to enable actionable improvements in LLM alignment, capability, and efficiency.
Activation patching provides evidence about neural network circuits when the choice of metric is aligned with the hypothesis and common interpretation errors are avoided.
Varying evaluation metrics and corruption methods in activation patching produces different localization and circuit discovery outcomes in language models, leading to recommendations for preferred practices.
citing papers explorer
-
WriteSAE: Sparse Autoencoders for Recurrent State
WriteSAE introduces sparse autoencoders with rank-1 matrix atoms for recurrent state updates, allowing replacement tests that outperform deletion on 92.4% of positions and a formula predicting logit changes with R²=0.98.
-
From Circuit Evidence to Mechanistic Theory: An Inductive Logic Approach
Introduces Causal Functional Signatures grounded in causal evidence and ILP-learned architectural signatures to enable explicit, comparable, and portable mechanistic claims across model scales.
-
Repeated-Token Counting Reveals a Dissociation Between Representations and Outputs
LLMs encode repeated token counts correctly in residual streams but a format-triggered MLP at 88-93% depth overwrites it with an incorrect fixed value.
-
In-Context Fixation: When Demonstrated Labels Override Semantics in Few-Shot Classification
In-context learning binds model outputs to the demonstrated label tokens as an exhaustive vocabulary, overriding semantic plausibility and causing fixation even with homogeneous or nonsense labels.
-
Navigating by Old Maps: The Pitfalls of Static Mechanistic Localization in LLM Post-Training
Transformer circuits show free evolution during SFT, rendering static mechanistic localization inadequate for future parameter updates due to inherent temporal latency.
-
The Linear Centroids Hypothesis: Features as Directions Learned by Local Experts
The Linear Centroids Hypothesis reframes network features as directions in centroid spaces of local affine experts, unifying interpretability methods and yielding sparser, more faithful dictionaries, circuits, and saliency maps.
-
Represented Is Not Computed: A Causal Test of Candidate Algorithmic Intermediates in a Transformer
Transformer represents but does not causally transmit staged algorithmic intermediates for base-digit extraction, diverging from probe predictions.
-
Dual-Pathway Circuits of Object Hallucination in Vision-Language Models
Vision-language models contain identifiable grounding and hallucination pathways; suppressing the latter reduces object hallucinations by up to 76% while preserving accuracy.
-
Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
-
When Reasoning Traces Become Performative: Step-Level Evidence that Chain-of-Thought Is an Imperfect Oversight Channel
CoT traces align with internal answer commitment in only 61.9% of steps on average, dominated by confabulated continuations after commitment has stabilized.
-
Instructions Shape Production of Language, not Processing
Instructions trigger a production-centered mechanism in language models, with task-specific information stable in input tokens but varying strongly in output tokens and correlating with behavior.
-
Patch-Effect Graph Kernels for LLM Interpretability
Patch-effect graphs built from causal mediation, partial correlation, and co-influence, when analyzed with graph kernels, preserve task-discriminative signals from activation patching that outperform global shape descriptors and raw baselines on GPT-2 Small.
-
Contrastive Attribution in the Wild: An Interpretability Analysis of LLM Failures on Realistic Benchmarks
Token-level contrastive attribution yields informative signals for some LLM benchmark failures but is not universally applicable across datasets and models.
-
Weight Patching: Toward Source-Level Mechanistic Localization in LLMs
Weight Patching localizes capabilities to specific parameter modules in LLMs by replacing weights from a behavior-specialized model into a base model and validating recovery via a vector-anchor interface, revealing a hierarchy of source, routing, and execution components.
-
Automated Attention Pattern Discovery at Scale in Large Language Models
AP-MAE reconstructs masked attention patterns in LLMs with high accuracy, generalizes across models, predicts generation correctness at 55-70%, and enables 13.6% accuracy gains via targeted interventions.
-
Reading Task Failure Off the Activations: A Sparse-Feature Audit of GPT-2 Small on Indirect Object Identification
An empirical audit identifies a strong SAE feature correlate for GPT-2 small failures on 'keys' prompts in the IOI task, performs ablation and baseline controls showing it is not causal, and presents the audit pipeline as the primary contribution.
-
Decodable but Not Corrected by Fixed Residual-Stream Linear Steering: Evidence from Medical LLM Failure Regimes
Overthinking in medical QA is linearly decodable at 71.6% accuracy yet fixed residual-stream steering yields no correction across 29 configurations, while enabling selective abstention with AUROC 0.610.
-
Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models
The survey organizes mechanistic interpretability techniques into a Locate-Steer-Improve framework to enable actionable improvements in LLM alignment, capability, and efficiency.
-
How to use and interpret activation patching
Activation patching provides evidence about neural network circuits when the choice of metric is aligned with the hypothesis and common interpretation errors are avoided.
-
Towards Best Practices of Activation Patching in Language Models: Metrics and Methods
Varying evaluation metrics and corruption methods in activation patching produces different localization and circuit discovery outcomes in language models, leading to recommendations for preferred practices.
- Where Does Authorship Signal Emerge in Encoder-Based Language Models?
- Language-Switching Triggers Take a Latent Detour Through Language Models
- High-Dimensional Statistics: Reflections on Progress and Open Problems
- How Language Models Process Negation