Spremljaj
Georg Lange
Naslov
Navedeno
Navedeno
Leto
Is this the subspace you are looking for? An interpretability illusion for subspace activation patching
A Makelov, G Lange, A Geiger, N Nanda
The Twelfth International Conference on Learning Representations, 2023
32023
An interpretability illusion for activation patching of arbitrary subspaces
G Lange, A Makelov, N Nanda
LessWrong, 2023
32023
Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Control
A Makelov, G Lange, N Nanda
arXiv preprint arXiv:2405.08366, 2024
2024
Quantifying Psychostimulant-induced Sensitization Effects on Dopamine and Acetylcholine Release across different Timescales
G Lange
2023
Reproducibility report for" Interpretable Complex-Valued Neural Networks for Privacy Protection"
A Sheverdin, N Corten, A Knijff, G Lange
ML Reproducibility Challenge 2020, 2021
2021
Sistem trenutno ne more izvesti postopka. Poskusite znova pozneje.
Članki 1–5