# Mechanistic Interpretability Domain: Artificial Intelligence · 234 papers indexed · newest 2026-10-28 Reverse-engineering neural networks: superposition, sparse autoencoders and transcoders, circuits and attribution graphs, probing and activation steering. Key concepts: mechanistic interpretability, sparse autoencoder, superposition, circuits, activation steering, probing, attribution graph, feature visualization ## Most influential - **[Systems biology and gene networks in neurodevelopmental and neurodegenerative disorders](https://space-kitty.com/papers/doi:10.1038/nrg3934)**. Neelroop Parikshak, Michael J. Gandal, Daniel H. Geschwind · 2015 · Nature Reviews Genetics (doi:10.1038/nrg3934 · 447 citations). Source: - **[Exploring single-cell data with deep multitasking neural networks](https://space-kitty.com/papers/doi:10.1038/s41592-019-0576-7)**. Matthew Amodio, David van Dijk, Krishnan Srinivasan et al. · 2019 · Nature Methods (doi:10.1038/s41592-019-0576-7 · 297 citations). Source: - **[Decoding of position in the developing neural tube from antiparallel morphogen gradients](https://space-kitty.com/papers/doi:10.1126/science.aam5887)**. Marcin Zagórski, Yoji Tabata, Nathalie Brandenberg et al. · 2017 · Science (doi:10.1126/science.aam5887 · 184 citations). Source: - **[Biologically-informed neural networks guide mechanistic modeling from sparse experimental data](https://space-kitty.com/papers/arxiv:2005.13073)**. John H. Lagergren, John T. Nardini, Ruth E. Baker et al. · 2020 · PLoS Computational Biology (arXiv:2005.13073 · doi:10.1371/journal.pcbi.1008462 · 167 citations). Source: > Biologically-informed neural networks (BINNs), an extension of physics-informed neural networks \[1\], are introduced and used to discover the underlying dynamics of biological systems from sparse experimental data. In the present work, BINNs are trained in a supervised learning framework to approximate in vitro cell biology assay experiments while respecting a generalized form of the governing reaction-diffusion partial differential equation (PDE). By allowing the diffusion and reaction terms to be multilayer perceptrons (MLPs), the nonlinear forms of these terms can be learned while simultaneo… - **[Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data](https://space-kitty.com/papers/doi:10.1186/s13059-020-02100-5)**. Nikolaus Fortelny, Christoph Bock · 2020 · Genome biology (doi:10.1186/s13059-020-02100-5 · 135 citations). Source: > BACKGROUND: Deep learning has emerged as a versatile approach for predicting complex biological phenomena. However, its utility for biological discovery has so far been limited, given that generic deep neural networks provide little insight into the biological mechanisms that underlie a successful prediction. Here we demonstrate deep learning on biological networks, where every node has a molecular equivalent, such as a protein or gene, and every edge has a mechanistic interpretation, such as a regulatory interaction along a signaling pathway. RESULTS: With knowledge-primed neural networks (KP… > _Abstract excerpt by Nikolaus Fortelny et al., CC-BY, https://doi.org/10.1186/s13059-020-02100-5; truncated._ - **[Machine learning for modeling N2O emissions from wastewater treatment plants: Aligning model performance, complexity, and interpretability](https://space-kitty.com/papers/doi:10.1016/j.watres.2023.120667)**. Mostafa Khalil, Ahmed AlSayed, Yang Liu et al. · 2023 · Water Research (doi:10.1016/j.watres.2023.120667 · 116 citations). Source: - **[Improving hydrologic models for predictions and process understanding using neural ODEs](https://space-kitty.com/papers/doi:10.5194/hess-26-5085-2022)**. Marvin Höge, Andreas Scheidegger, Marco Baity‐Jesi et al. · 2022 · Hydrology and earth system sciences (doi:10.5194/hess-26-5085-2022 · 126 citations). Source: > Abstract. Deep learning methods have frequently outperformed conceptual hydrologic models in rainfall-runoff modelling. Attempts of investigating the internals of such deep learning models are being made but traceability of model states and processes and their interrelations to model input and output is not yet fully given. Direct interpretability of mechanistic processes has always been considered as asset of conceptual models that helps to gain system understanding aside of predictability. We introduce hydrologic Neural Ordinary Differential Equation (ODE) models that perform as well as stat… > _Abstract excerpt by Marvin Höge et al., CC-BY, https://doi.org/10.5194/hess-26-5085-2022; truncated._ - **[Improving accuracy of rutting prediction for mechanistic-empirical pavement design guide with deep neural networks](https://space-kitty.com/papers/doi:10.1016/j.conbuildmat.2018.09.087)**. Hongren Gong, Yiren Sun, Zijun Mei et al. · 2018 · Construction and Building Materials (doi:10.1016/j.conbuildmat.2018.09.087 · 130 citations). Source: - **[Generative Retrieval-Augmented Ontologic Graph and Multiagent Strategies for Interpretive Large Language Model-Based Materials Design](https://space-kitty.com/papers/arxiv:2310.19998)**. Markus J. Buehler · 2024 · ACS Engineering Au (arXiv:2310.19998 · doi:10.1021/acsengineeringau.3c00058 · 82 citations). Source: > Transformer neural networks show promising capabilities, in particular for uses in materials analysis, design and manufacturing, including their capacity to work effectively with both human language, symbols, code, and numerical data. Here we explore the use of large language models (LLMs) as a tool that can support engineering analysis of materials, applied to retrieving key information about subject areas, developing research hypotheses, discovery of mechanistic relationships across disparate areas of knowledge, and writing and executing simulation codes for active knowledge generation based… - **[Mechanistic Differences in the Transcriptional Interpretation of Local and Long-Range Shh Morphogen Signaling](https://space-kitty.com/papers/doi:10.1016/j.devcel.2012.09.015)**. Tony Oosterveen, Sanja Kurdija, Zhanna Alekseenko et al. · 2012 · Developmental Cell (doi:10.1016/j.devcel.2012.09.015 · 132 citations). Source: - **[Development of a hybrid model for a partially known intracellular signaling pathway through correction term estimation and neural network modeling](https://space-kitty.com/papers/doi:10.1371/journal.pcbi.1008472)**. Dongheon Lee, Arul Jayaraman, Joseph Sang‐Il Kwon · 2020 · PLoS Computational Biology (doi:10.1371/journal.pcbi.1008472 · 103 citations). Source: > Developing an accurate first-principle model is an important step in employing systems biology approaches to analyze an intracellular signaling pathway. However, an accurate first-principle model is difficult to be developed since it requires in-depth mechanistic understandings of the signaling pathway. Since underlying mechanisms such as the reaction network structure are not fully understood, significant discrepancy exists between predicted and actual signaling dynamics. Motivated by these considerations, this work proposes a hybrid modeling approach that combines a first-principle model and… > _Abstract excerpt by Dongheon Lee et al., CC-BY, https://doi.org/10.1371/journal.pcbi.1008472; truncated._ - **[A review of mechanistic learning in mathematical oncology](https://space-kitty.com/papers/doi:10.3389/fimmu.2024.1363144)**. John Metzcar, Catherine Ruth Jutzeler, Paul Macklin et al. · 2024 · Frontiers in Immunology (doi:10.3389/fimmu.2024.1363144 · 70 citations). Source: > Mechanistic learning refers to the synergistic combination of mechanistic mathematical modeling and data-driven machine or deep learning. This emerging field finds increasing applications in (mathematical) oncology. This review aims to capture the current state of the field and provides a perspective on how mechanistic learning may progress in the oncology domain. We highlight the synergistic potential of mechanistic learning and point out similarities and differences between purely data-driven and mechanistic approaches concerning model complexity, data requirements, outputs generated, and in… > _Abstract excerpt by John Metzcar et al., CC-BY, https://doi.org/10.3389/fimmu.2024.1363144; truncated._ - **[Efficient inverse graphics in biological face processing](https://space-kitty.com/papers/doi:10.1126/sciadv.aax5979)**. Ilker Yildirim, Mario Belledonne, Winrich A. Freiwald et al. · 2020 · Science Advances (doi:10.1126/sciadv.aax5979 · 89 citations). Source: - **[Explainable AI: A review of applications to neuroimaging data](https://space-kitty.com/papers/doi:10.3389/fnins.2022.906290)**. Farzad V. Farahani, Krzysztof Fiok, Behshad Lahijanian et al. · 2022 · Frontiers in Neuroscience (doi:10.3389/fnins.2022.906290 · 81 citations). Source: > Deep neural networks (DNNs) have transformed the field of computer vision and currently constitute some of the best models for representations learned via hierarchical processing in the human brain. In medical imaging, these models have shown human-level performance and even higher in the early diagnosis of a wide range of diseases. However, the goal is often not only to accurately predict group membership or diagnose but also to provide explanations that support the model decision in a context that a human can readily interpret. The limited transparency has hindered the adoption of DNN algori… > _Abstract excerpt by Farzad V. Farahani et al., CC-BY, https://doi.org/10.3389/fnins.2022.906290; truncated._ - **[ToxAIcology - The evolving role of artificial intelligence in advancing toxicology and modernizing regulatory science](https://space-kitty.com/papers/doi:10.14573/altex.2309191)**. Thomas Härtung · 2023 · ALTEX (doi:10.14573/altex.2309191 · 77 citations). Source: > Toxicology has undergone a transformation from an observational science to a data-rich discipline ripe for artificial intelligence (AI) integration. The exponential growth in computing power coupled with accumulation of large toxicological datasets has created new opportunities to apply techniques like machine learning and especially deep learning to enhance chemical hazard assessment. This article provides an overview of key developments in AI-enabled toxicology, including early expert systems, statistical learning methods like quantitative structure-activity relationships (QSARs), recent adv… > _Abstract excerpt by Thomas Härtung, CC-BY, https://doi.org/10.14573/altex.2309191; truncated._ ## Newest - **[The Prestige: Benchmarking Cognitive Visual Reasoning using Magic Tricks](https://space-kitty.com/papers/arxiv:2610.38103)**. Shuo Wen, Beixi Du, Edwin Meriaux et al. · 2026 (arXiv:2610.38103). Source: - **[Replaying the Same Gradients at a Different $\\varepsilon$: A Registered Test of the Fixed-Trajectory Account of Adam's $\\varepsilon$ for One Rare Row](https://space-kitty.com/papers/arxiv:2610.38169)**. Lngel Kyo · 2026 (arXiv:2610.38169). Source: - **[Scaling Lean Theorem Proving for Real-World Code Verification](https://space-kitty.com/papers/arxiv:2610.38174)**. Debargha Ganguly, Divya Raghunathan, Dejan Jovanović et al. · 2026 (arXiv:2610.38174). Source: - **[Barren Plateaus in Quantum Linear Combinations: Interference and Lie-Algebraic Control](https://space-kitty.com/papers/arxiv:2610.38193)**. Amir Koutahi, Pouya Kananian, Hans-Arno Jacobsen · 2026 (arXiv:2610.38193). Source: - **[From Natural Language to Real Schematics: An Automated Workflow for Power Electronics Schematic Synthesis](https://space-kitty.com/papers/arxiv:2610.38201)**. Bowen Su, Huan Chen, Jiahua Ying et al. · 2026 (arXiv:2610.38201). Source: - **[Robust LLM-Generated Text Detection through Matched Conditional-Response Probing](https://space-kitty.com/papers/arxiv:2610.38312)**. Junfeng Hao, Wenxian Wang, Tao Leng et al. · 2026 (arXiv:2610.38312). Source: - **[Can LLMs Faithfully Solve Ontology Engineering Tasks? A Mechanistic Interpretability Study](https://space-kitty.com/papers/arxiv:2610.38326)**. Bohui Zhang, Albert Meroño-Peñuela, Elena Simperl · 2026 (arXiv:2610.38326). Source: - **[SpaceAct: Benchmarking 3D Spatial and Temporal Reasoning via Closed-Loop Action and Active Perception](https://space-kitty.com/papers/arxiv:2610.38405)**. Chenglong Shi, Zekai Yuan, Yue Song et al. · 2026 (arXiv:2610.38405). Source: - **[What Hallucinations Have We Measured? Semantic Illusion, Confabulation, and Expressed Uncertainty](https://space-kitty.com/papers/arxiv:2610.38520)**. Hazel H. Kim, Jakub Binkowski, Ekaterina Fadeeva et al. · 2026 (arXiv:2610.38520). Source: - **[Value Probes: Learned Value Functions for Beam-Search Red Teaming](https://space-kitty.com/papers/arxiv:2610.38526)**. Jonathan F. Jao, Andrew Kyle Lampinen · 2026 (arXiv:2610.38526). Source: --- Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.