Model Releases
ADAG: Automatically Describing Attribution Graphs
arXiv:2604.07615v1 Announce Type: new Abstract: In language model interpretability research, extbf{circuit tracing} aims to identify which internal features causally contributed to a particular outp
arXiv:2604.07615v1 Announce Type: new Abstract: In language model interpretability research, extbf{circuit tracing} aims to identify which internal features causally contributed to a particular output and how they affected each other, with the goal of explaining the computations underlying some behaviour. However, all prior circuit tracing work has relied on ad-hoc human interpretation of the role that each feature in the circuit plays, via manual inspection of data artifacts such as the dataset examples the component activates on. We introduce extbf{ADAG}, an end-to-end pipeline for describing these attribution graphs which is fully automated. To achieve this, we introduce extit{attribution profiles} which quantify the functional role of a feature via its input and output gradient effects. We then introduce a novel clustering algorithm for grouping features, and an LLM explainer--simulator setup which generates and scores natural-language explanations of the functional role of these feature groups. We run our system on known human-analysed circuit-tracing tasks and recover interpretable circuits, and further show ADAG can find steerable clusters which are responsible for a harmful advice jailbreak in Llama 3.1 8B Instruct.
Related
- Emotion Concepts and their Function in a Large Language Model
- BenchBrowser: Retrieving Evidence for Evaluating Benchmark Validity
- ModeX: Evaluator-Free Best-of-N Selection for Open-Ended Generation
- Beyond Social Pressure: Benchmarking Epistemic Attack in Large Language Models
Source: arXiv cs.CL | 2026-04-10