5 papers
Auditing Games for Sandbagging
Jordan Taylor, Sid Black, Dillon Bowen +10
Future AI systems could conceal their capabilities ('sandbagging') during evaluations, potentially misleading developers and auditors. We stress-tested sandbagging detection techni…
Evaluating Explanations: An Explanatory Virtues Framework for Mechanistic Interpretability -- The Strange Science Part I.ii
Kola Ayonrinde, Louis Jaburi
Mechanistic Interpretability (MI) aims to understand neural networks through causal explanations. Though MI has many explanation-generating methods, progress has been limited by th…
A Mathematical Philosophy of Explanations in Mechanistic Interpretability -- The Strange Science Part I.i
Kola Ayonrinde, Louis Jaburi
Mechanistic Interpretability aims to understand neural networks through causal explanations. We argue for the Explanatory View Hypothesis: that Mechanistic Interpretability researc…
SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability
Adam Karvonen, Can Rager, Johnny Lin +12
Sparse autoencoders (SAEs) are a popular technique for interpreting language model activations, and there is extensive recent work on improving SAE effectiveness. However, most pri…
Adaptive Sparse Allocation with Mutual Choice & Feature Choice Sparse Autoencoders
Kola Ayonrinde
Sparse autoencoders (SAEs) are a promising approach to extracting features from neural networks, enabling model interpretability as well as causal interventions on model internals.…