LESSWRONG
LW

To anyone reading this who wants to work on or discuss FHI-flavored work: Consider applying to Constellation's programs (the deadline for some of them is today!), which include salaried positions for researchers.

Apply to the Constellation Visiting Researcher Program and Astra Fellowship, in Berkeley this Winter

Nate Thomas2yΩ120

Thanks, Neel! It should be fixed now.

Takeaways from our robust injury classifier project [Redwood Research]

Nate Thomas3yΩ91812

Note that it's unsurprising that a different model categorizes this correctly because the failure was generated from an attack on the particular model we were working with. The relevant question is "given a model, how easy is it to find a failure by attacking that model using our rewriting tools?"

42Apply to the Constellation Visiting Researcher Program and Astra Fellowship, in Berkeley this Winter

34Causal scrubbing: results on induction heads

34Causal scrubbing: results on a paren balance checker

18Causal scrubbing: Appendix

206Causal Scrubbing: a method for rigorously testing interpretability hypotheses [Redwood Research]

135Apply to the Redwood Research Mechanistic Interpretability Experiment (REMIX), a research program in Berkeley

142High-stakes alignment via adversarial training [Redwood Research report]

56We're Redwood Research, we do applied alignment research, AMA