Bridging the Black Box: A Survey on Mechanistic Interpretability in AI
ACM Computing Surveys · 30.06.2026 · 58 (8) · 1-35
DOI: 10.1145/3787104 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Computing Surveys, online erschienen am 04.02.2026. Verfasst von Shriyank Somvanshi und 9 weiteren Autor:innen. Die Arbeit wurde bisher 15-mal zitiert (Crossref) und nennt 128 Literaturangaben.
Abstract (Originalsprache)
Mechanistic interpretability seeks to reverse-engineer the internal logic of neural networks by uncovering human-understandable circuits, algorithms, and causal structures that drive model behavior. Unlike post hoc explanations that describe what models do, this paradigm focuses on why and how they compute, tracing information flow through neurons, attention heads, and activation pathways. This survey provides a high-level synthesis of the field-highlighting its motivation, conceptual foundations, and methodological taxonomy rather than enumerating individual techniques. We organize mechanistic interpretability across three abstraction layers— neurons , circuits , and algorithms —and three evaluation perspectives: behavioral , counterfactual , and causal . We further discuss representative approaches and toolchains that enable structural analysis of modern AI systems, outlining how mechanistic interpretability bridges theoretical insights with practical transparency. Despite rapid progress, challenges persist in scaling these analyses to frontier models, resolving polysemantic representations, and establishing standardized causal benchmarks. By connecting historical evolution, current methodologies, and emerging research directions, this survey aims to provide an integrative framework for understanding how mechanistic interpretability can support transparency, reliability, and governance in large-scale AI.
Zitieren
APA
Somvanshi, S., Islam, M. M., Rafe, A., Tusti, A. G., Chakraborty, A., Baitullah, A., Chowdhury, T. I., Alnawmasi, N., Dutta, A., & Das, S. (2026). Bridging the Black Box: A Survey on Mechanistic Interpretability in AI. ACM Computing Surveys, 58(8), 1-35. https://doi.org/10.1145/3787104
Vancouver
Somvanshi S, Islam MM, Rafe A, Tusti AG, Chakraborty A, Baitullah A, et al. Bridging the Black Box: A Survey on Mechanistic Interpretability in AI. ACM Computing Surveys. 2026;58(8):1-35. doi:10.1145/3787104
BibTeX
@article{Somvanshi2026,
author = {Somvanshi, Shriyank and Islam, Md Monzurul and Rafe, Amir and Tusti, Anannya Ghosh and Chakraborty, Arka and Baitullah, Anika and Chowdhury, Tausif Islam and Alnawmasi, Nawaf and Dutta, Anandi and Das, Subasish},
title = {Bridging the Black Box: A Survey on Mechanistic Interpretability in AI},
journal = {ACM Computing Surveys},
year = {2026},
volume = {58},
number = {8},
pages = {1--35},
doi = {10.1145/3787104}
}
EUVIMED