# Circuit-Diff: Factual Edit-based Intervention Method for Localizing Knowledge in Attribution Graphs

arXiv: https://arxiv.org/abs/2609.23892 · HTML: https://mlsanity.com/paper/2609.23892/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability.
- Builds on 19 of its 30 references that are in the map.
- First posted on arXiv 20 Sep 2026.

## Placed in

- Mechanistic interpretability / [Circuits and mechanistic analysis](https://mlsanity.com/mechinterp/circuits-and-mechanistic-analysis/)

## Builds on (most cited first, 19 of 19)

- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Mass-Editing Memory in a Transformer](https://mlsanity.com/paper/2210.07229/) (2022)
- [Toy Models of Superposition](https://mlsanity.com/paper/2209.10652/) (2022)
- [Dissecting Recall of Factual Associations in Auto-Regressive Language Models](https://mlsanity.com/paper/2304.14767/) (2023)
- [Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Models](https://mlsanity.com/paper/2301.04213/) (2023)
- [Transcoders Find Interpretable LLM Feature Circuits](https://mlsanity.com/paper/2406.11944/) (2024)
- [Linear Algebraic Structure of Word Senses, with Applications to Polysemy](https://mlsanity.com/paper/1601.03764/) (2016)
- [Automatically Interpreting Millions of Features in Large Language Models](https://mlsanity.com/paper/2410.13928/) (2024)
- [Understanding Transformer Memorization Recall Through Idioms](https://mlsanity.com/paper/2210.03588/) (2022)
- [Detecting Edit Failures In Large Language Models: An Improved Specificity Benchmark](https://mlsanity.com/paper/2305.17553/) (2023)
- [Understanding Factual Recall in Transformers via Associative Memories](https://mlsanity.com/paper/2412.06538/) (2024)
- [Rigorously Assessing Natural Language Explanations of Neurons](https://mlsanity.com/paper/2309.10312/) (2023)
- [Long-form evaluation of model editing](https://mlsanity.com/paper/2402.09394/) (2024)
- [Circuit Tracing in Vision-Language Models: Understanding the Internal Mechanisms of Multimodal Thinking](https://mlsanity.com/paper/2602.20330/) (2026)
- [Understanding the Collapse of LLMs in Model Editing](https://mlsanity.com/paper/2406.11263/) (2024)
- [DifFRACT: Diffusion Feature Reconstruction and Attribution for Circuit Tracing](https://mlsanity.com/paper/2606.15796/) (2026)
- [Automated Attribution Graph Interpretation via Probe Prompting](https://mlsanity.com/paper/2511.07002/) (2025)
- [CRaFT: Circuit-Guided Refusal Feature Selection via Cross-Layer Transcoders](https://mlsanity.com/paper/2604.01604/) (2026)
- [Prune, Interpret, Evaluate: A Cross-Layer Transcoder-Native Framework for Efficient Circuit Discovery via Feature Attribution](https://mlsanity.com/paper/2604.16889/) (2026)

## Abstract (arXiv)

Mechanistic interpretability defines features as the fundamental units of a neural network and circuits as the weighted subgraphs that carry out its computation. Because individual neurons are polysemantic, Cross-Layer Transcoders (CLTs) were introduced as a way to approximate a model's circuits by generating an attribution graph. The nodes of that graph, however, are unlabeled features: reading a graph means pruning it and then working out by hand what each surviving node means. To make CLTs easier to use for circuit discovery, we introduce Circuit-Diff, which intervenes on the model itself with a low-rank factual edit and takes the features whose role in the attribution graph changes under that edit as related to the edited knowledge. On the edits we examine, the flagged nodes are not only detectors of the object token: read off the CLT's released feature dashboards, they include features for the history, geography and associations surrounding the old and new objects. We formalize the method, measure how reliable a frozen CLT remains after a factual edit, test the selected nodes causally by patching them on up to 24 CounterFact edits, give a case study, and release an open-source implementation built on the circuit-tracer package, together with two further tools (multi-prompt aggregation and rule-based supernode labeling).