Neel Nanda


GDM Mech Interp Progress Updates
Fact Finding: Attempting to Reverse-Engineer Factual Recall on the Neuron Level
Interpreting Othello-GPT
200 Concrete Open Problems in Mechanistic Interpretability

Wiki Contributions


I found this comment very helpful, and also expected probing to be about as good, thank you!

Glad you liked the post!

I'm also pretty interested in combining steering vectors. I think a particularly promising direction is using SAE decoder vectors for this, as SAEs are designed to find feature vectors that independently vary and can be added.

I agree steering vectors are important as evidence for the linear representation hypothesis (though at this point I consider SAEs to be much superior as evidence, and think they're more interesting to focus on)

I'm not aware of any problems with it. I think it's a nice paper, but not really at my bar for important work (which is a really high bar, to be clear - fewer than half the papers in this post probably meet it)

Fair point, I'll add that in to the post. The main reason I recommend it so highly and prominently is that I think it builds valuable conceptual frameworks for reasoning about the pieces of a transformer, even if it somewhat overclaims on how far it can get on interpreting tiny attention-only models, and I think those broad intuitions still stand even after your critiques. Eg strict induction heads as an example of the kind of algorithm that can be implemented with attention, even if it's not fully faithful to the underlying model. But I agree that these are worthwhile caveats to have in mind when reading, and the paper shouldn't be blindly recommended.

This is a preference model trained on GPT-J I think, so my guess is that it's just very dumb and learned lots of silly features. I'd be very surprise if a ChatGPT preference model had the same issues when an SAE is trained on it.

This was interesting, thanks! I really enjoy your short book reviews

I think it was an interesting paper, but this analysis and predictions all seem extremely on point to me

Oh, that's great! Kudos to the authors for setting the record straight. I'm glad your work is now appropriately credited

Thanks! Note that this work uses steering vectors, not SAEs, so the technique is actually really easy and cheap - I actively think this is one of the main selling points (you can jailbreak a 70B model in minutes, without any finetuning or optimisation). I am excited at the idea of seeing if you can improve it with SAEs though - it's not obvious to me that SAEs are better than steering vectors, though it's plausible.

I may take you up on the two hours offer, thanks! I'll ask my co-authors

But it mostly seems like it would be helpful because it gives you well-tuned baselines to compare your results to. I don't think you have results that can cleanly be compared to well-established baselines?

If we compared our jailbreak technique to other jailbreaks on an existing benchmark like Harm Bench and it does comparably well to SOTA techniques, or does even better than SOTA techniques, would you consider this success at doing something useful on a real task?

Load More