Profile avatar
meatlearner.bsky.social
Machine-learner, meat-learner, research scientist, AI Safety thinker. Model trainer, skeptical adorer of statistics. Co-author of: Malware Data Science
11 posts 39 followers 117 following
Posts 4 Comments 7

In honor of MLK day, here's super interesting essay my partner wrote on Martin Luther King Jr: what he actually believed and accomplished (different than what is sometimes described). docs.google.com/document/d/1... Incredibly impressive person.

submitted 42 days ago • 0 comments

Sleeper Agents arxiv.org/pdf/2401.05566 So many AI safety issues get worse, & harder to combat the larger and more advanced your model gets: "The backdoor behavior is most persistent in the largest models and in models trained to produce chain-of-thought reasoning"

submitted 90 days ago • 1 comment

Anthropic's "Towards Sycophancy In Language Models" arxiv.org/pdf/2310.13548 TLDR: LLMs tend to generate sycophantic responses. Human feedback & preference models encourage this behavior. I also think this is just the nature of training on internet writing.... We write in social clusters:

submitted 90 days ago • 1 comment

In AI safety, we have inner misalignment (actions don't minimize the loss function) and outer misalignment (loss function is misspecified). But I do think that inner misalignment (~learned features) tend to act as a protective mechanism to avoid outer misalignment implications. I, er, really hope.

submitted 100 days ago • 1 comment