171–175 of 175 posts
Earlier quoted context omitted.
Then why do Chess AI perform much better than LLMs trying to play chess.
Because they have an explicit model of chess and specific heuristics for learning chess. An LLM could have picked up some chess patterns through osmosis, but it can not reason explicitly in the domain.
Earlier quoted context omitted.
Then why do Chess AI perform much better than LLMs trying to play chess.
Why would a language model be good at playing chess?
Earlier quoted context omitted.
They're probably referring to fine-tuning on private/proprietary data that is specific to a use case. Say a history of conversation transcripts in a call center. Larger models, like OpenAI's GPT, don't have access to this by default.
OpenAI’s API has fine tuning options for older GPT models: davinci, curie, babbage, and ada
Earlier quoted context omitted.
Thanks for the great project! Any chance, your team might consider more open platform than Discord for posting updates? I personally find Discord hard to use, and there’s no way to have sensible subscription (like RSS). Discord is usually muted.
Discord is a black hole where information goes to die. Its search and scrollback is awful. It's awful at being an archive, as finding anything that was asked more than a day or two ago is impractical. To use Discord in good faith and with open eyes, you have to prioritize communication in the present, and give up hope of archiving anything that was said for people who might need the information in the future.
Its a shame its being used for general communities who don't use or need the voice chat feature. Especially when its an official community for a place, given their stance on third party clients and privacy issues.
If you don't need voice chat, Zulip, Mattermost, Revolt, Discourse, and many other would suffice (Linen recently got featered on HN). If you do, I think even Signal would be suffice these days.
For Discord search, recently Answer Overflow was recently featured on HN [1].
Earlier quoted context omitted.
I’m not sure ML researchers would agree that number of (compressed) bytes are more meaningful than number of parameters. Parameters have mathematical meaning – bytes doesn’t.
My point is that they don't have a mathematical meaning. They have a training-time impact, but that's more relevant when creating than using. You'd need to know how many parameters were independently covarying for any given class of predictions. It certainly isnt all of them. You could cite the "average dropout percent to random-level accuracy on a given class of problems" (my guess is that this would show 5-20% of p…