logoalt Hacker News

tb99 • today at 6:52 PM • 2 replies • view on HN

It's also interesting to look at how the agent does the fix. Appendix E5 shows what it's doing:

- Find a common category of error (too many tokens).

- Calculate aggregated difference between states in examples in that error category and other examples.

- Update weights to push states along that difference direction toward the non-error states.

In other words, deprived of gradient descent as a tool, it approximates it. It's clever, and it's a neat experiment. It doesn't show any superhuman insight into its weights -- with the same constraints, a human would likely take a similar approach. But it's early days for self modification and it will be interesting to see if models get a deeper understanding of weights as they get smarter.


Replies

cavoirom • today at 7:06 PM

I think they will get deeper understanding of weights as future human research will be in the training data.