Maxime Peyrard @peyrardmax.bsky.social · 21/04/2025In our work, we stress-test the identifiability of research programs of MI with small MLPs and simple boolean logic tasks. Why? It allows us to enumerate all possible explanations and see how many pass various MI testing criteria. 100
Maxime Peyrard @peyrardmax.bsky.social · 21/04/2025We find a lot of identifiability issues: - Multiple explanatory algorithms exists - Even for one algorithm, there are many localizations in the network Identifiability problems remain across scenarios: changing levels of over-parametrization, progress in training, multi-tasks, model size. 110
Maxime Peyrard @peyrardmax.bsky.social · 21/04/2025What can be done? 👉 Stricter validity criteria? 👉 Maybe interpretability is inherently underdetermined? and we can only get control and predictability but not "understanding" This is a fascinating topic, and we keep investigating. If you're interested, come and chat at ICLR! 010
