# Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligence arXiv: https://arxiv.org/abs/2608.12036 ยท HTML: https://mlsanity.com/paper/2608.12036/ ## Key facts - Cited by 2 papers in the map (data through 14 Oct 2026). - Builds on 72 of its 101 references that are in the map. - First posted on arXiv 12 Aug 2026. ## Builds on (most cited first, 30 of 72) - [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017) - [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022) - [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023) - [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020) - [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023) - [Editing Models with Task Arithmetic](https://mlsanity.com/paper/2212.04089/) (2022) - [Steering Llama 2 via Contrastive Activation Addition](https://mlsanity.com/paper/2312.06681/) (2023) - [Towards Automated Circuit Discovery for Mechanistic Interpretability](https://mlsanity.com/paper/2304.14997/) (2023) - [Eliciting Latent Predictions from Transformers with the Tuned Lens](https://mlsanity.com/paper/2303.08112/) (2023) - [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024) - [Dissecting Recall of Factual Associations in Auto-Regressive Language Models](https://mlsanity.com/paper/2304.14767/) (2023) - [MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering](https://mlsanity.com/paper/2410.07095/) (2024) - [AFlow: Automating Agentic Workflow Generation](https://mlsanity.com/paper/2410.10762/) (2024) - [Mechanistic Interpretability for AI Safety -- A Review](https://mlsanity.com/paper/2404.14082/) (2024) - [Accelerating scientific discovery with Co-Scientist](https://mlsanity.com/paper/2502.18864/) (2025) - [Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2](https://mlsanity.com/paper/2408.05147/) (2024) - [Alignment faking in large language models](https://mlsanity.com/paper/2412.14093/) (2024) - [Consistent Individualized Feature Attribution for Tree Ensembles](https://mlsanity.com/paper/1802.03888/) (2018) - [LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations](https://mlsanity.com/paper/2410.02707/) (2024) - [Attribution Patching Outperforms Automated Circuit Discovery](https://mlsanity.com/paper/2310.10348/) (2023) - [Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge](https://mlsanity.com/paper/2407.19594/) (2024) - [Transcoders Find Interpretable LLM Feature Circuits](https://mlsanity.com/paper/2406.11944/) (2024) - [OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts](https://mlsanity.com/paper/2205.01833/) (2022) - [Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs](https://mlsanity.com/paper/2502.17424/) (2025) - [A Comprehensive Study of Knowledge Editing for Large Language Models](https://mlsanity.com/paper/2401.01286/) (2024) - [A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders](https://mlsanity.com/paper/2409.14507/) (2024) - [Towards End-to-End Automation of AI Research](https://mlsanity.com/paper/2606.15497/) (2026) - [Have Faith in Faithfulness: Going Beyond Circuit Overlap When Finding Model Mechanisms](https://mlsanity.com/paper/2403.17806/) (2024) - [Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoders](https://mlsanity.com/paper/2410.20526/) (2024) - [From Attribution Maps to Human-Understandable Explanations through Concept Relevance Propagation](https://mlsanity.com/paper/2206.03208/) (2022) ## Built on by (most cited first, 2 of 2) - [ConflictGuide: AutoResearch Improves When Competing Behaviors Are Made Visible](https://mlsanity.com/paper/2609.39933/) (2026) - [EurekaBench: Measuring Agentic Ability to Discover New Scientific Insights](https://mlsanity.com/paper/2610.00492/) (2026) ## Abstract (arXiv) AI models are increasingly used in scientific discovery and human decision-making. Yet how AI models work and what risks they pose remain poorly understood. As AI development becomes faster and more automated, research on the mechanisms underlying AI remains largely manual. To bridge this gap, we introduce Mechanist, an agentic system that uses AI as a scientific instrument for the autonomous discovery of mechanisms underlying AI. To ground novel mechanism hypotheses, we construct a scientific knowledge graph of 13,000 studies on AI mechanisms, alongside a multidisciplinary database of 43 million papers spanning 26 fields. For reliable experiment execution, we curate a library of 32 foundational methods for mechanism analysis. Compared with Claude Code and existing AI-scientist systems, Mechanist generates higher-quality mechanism hypotheses and executes experiments more reliably. Across four case studies, Mechanist autonomously discovers new model behaviors and their underlying mechanisms, and translates these discoveries into mechanism-guided interventions and interdisciplinary design. Specifically, Mechanist first uncovers a counterintuitive safety risk in scientific laboratories, showing that unsafe traits can transfer to fine-tuned student models through apparently safe training data and emerge across modalities. Mechanist then develops a mechanism theory of belief, revealing how models represent world knowledge, form beliefs, infer the beliefs of others, and how these mechanisms emerge during pretraining. Building on this theory, Mechanist develops targeted interventions that improve model performance across diverse scenarios. Finally, Mechanist can also advance interdisciplinary discovery through mechanistic design, providing an alternative to the computationally intensive generate-and-rerank paradigm.