Cover Image

Mechanistic Interpretability for AI Safety -- A Review

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Bereska, Leonard, Gavves, Efstratios
Format:	Preprint
Published:	2024
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Superposition as Lossy Compression: Measure with Sparse Autoencoders and Connect to Adversarial Vulnerability
by: Bereska, Leonard, et al.
Published: (2025)

Mechanistic Neural Networks for Scientific Machine Learning
by: Pervez, Adeel, et al.
Published: (2024)

From MLP to NeoMLP: Leveraging Self-Attention for Neural Fields
by: Kofinas, Miltiadis, et al.
Published: (2024)

Physics-Guided Radiotherapy Treatment Planning with Deep Learning
by: Achlatis, Stefanos, et al.
Published: (2025)

Unleashing Uncertainty: Efficient Machine Unlearning for Generative AI
by: Spartalis, Christoforos N., et al.
Published: (2025)

From Explainable to Explained AI: Ideas for Falsifying and Quantifying Explanations
by: Schirris, Yoni, et al.
Published: (2025)

Language Agents Meet Causality -- Bridging LLMs and Causal World Models
by: Gkountouras, John, et al.
Published: (2024)

LoTUS: Large-Scale Machine Unlearning with a Taste of Uncertainty
by: Spartalis, Christoforos N., et al.
Published: (2025)

Mechanistic PDE Networks for Discovery of Governing Equations
by: Pervez, Adeel, et al.
Published: (2025)

CTRL-O: Language-Controllable Object-Centric Visual Representation Learning
by: Didolkar, Aniket, et al.
Published: (2025)

Distributional Vision-Language Alignment by Cauchy-Schwarz Divergence
by: Yin, Wenzhe, et al.
Published: (2025)

Towards Unified Attribution in Explainable AI, Data-Centric AI, and Mechanistic Interpretability
by: Zhang, Shichang, et al.
Published: (2025)

AI Alignment Strategies from a Risk Perspective: Independent Safety Mechanisms or Shared Failures?
by: Dung, Leonard, et al.
Published: (2025)

Mechanistic Interpretability of LoRA-Adapted Language Models for Nuclear Reactor Safety Applications
by: Lee, Yoon Pyo
Published: (2025)

Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability
by: Geiger, Atticus, et al.
Published: (2023)

Mechanistic Interpretability Needs Philosophy
by: Williams, Iwan, et al.
Published: (2025)

Mechanistically Interpreting Compression in Vision-Language Models
by: Elluru, Veeraraju, et al.
Published: (2026)

Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods
by: Grey, Markov, et al.
Published: (2025)

A Practical Review of Mechanistic Interpretability for Transformer-Based Language Models
by: Rai, Daking, et al.
Published: (2024)

Challenges in Mechanistically Interpreting Model Representations
by: Golechha, Satvik, et al.
Published: (2024)

Mechanistic Interpretability in the Presence of Architectural Obfuscation
by: Florencio, Marcos, et al.
Published: (2025)

Neural Transparency: Mechanistic Interpretability Interfaces for Anticipating Model Behaviors for Personalized AI
by: Karny, Sheer, et al.
Published: (2025)

MIB: A Mechanistic Interpretability Benchmark
by: Mueller, Aaron, et al.
Published: (2025)

nnterp: A Standardized Interface for Mechanistic Interpretability of Transformers
by: Dumas, Clément
Published: (2025)

Dissecting Bias in LLMs: A Mechanistic Interpretability Perspective
by: Chandna, Bhavik, et al.
Published: (2025)

On the Mechanistic Interpretability of Neural Networks for Causality in Bio-statistics
by: Conan, Jean-Baptiste A.
Published: (2025)

MINAR: Mechanistic Interpretability for Neural Algorithmic Reasoning
by: He, Jesse, et al.
Published: (2026)

Graph Neural Networks for Learning Equivariant Representations of Neural Networks
by: Kofinas, Miltiadis, et al.
Published: (2024)

reward-lens: A Mechanistic Interpretability Library for Reward Models
by: Nadaf, Mohammed Suhail B
Published: (2026)

A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models
by: Lin, Zihao, et al.
Published: (2025)

International Agreements on AI Safety: Review and Recommendations for a Conditional AI Safety Treaty
by: Scholefield, Rebecca, et al.
Published: (2025)

Mechanistic Interpretability of Emotion Inference in Large Language Models
by: Tak, Ala N., et al.
Published: (2025)

Faithful and Stable Neuron Explanations for Trustworthy Mechanistic Interpretability
by: Yan, Ge, et al.
Published: (2025)

Mechanistic Interpretability for Transformer-based Time Series Classification
by: Kalnāre, Matīss, et al.
Published: (2025)

From Demonstrations to Rewards: Test-Time Prompt Optimization for VLM Reward Models
by: Gumbsch, Christian, et al.
Published: (2026)

Grounding Continuous Representations in Geometry: Equivariant Neural Fields
by: Wessels, David R, et al.
Published: (2024)

Mechanistic Interpretability as Statistical Estimation: A Variance Analysis
by: Méloux, Maxime, et al.
Published: (2025)

Space-Time Continuous PDE Forecasting using Equivariant Neural Fields
by: Knigge, David M., et al.
Published: (2024)

Towards Understanding Safety Alignment: A Mechanistic Perspective from Safety Neurons
by: Chen, Jianhui, et al.
Published: (2024)

Mechanistic Interpretability with Sparse Autoencoder Neural Operators
by: Tolooshams, Bahareh, et al.
Published: (2025)