Talking About Large Language Models · HackerPapers discussions

Talking About Large Language Models — HN submission

Sat, 08 Jun 2024 01:05:46 GMT

Talking About Large Language Models (LLMs)

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Talking About Large Language Models — HN submission

Mon, 23 Jan 2023 13:58:30 GMT

Talking About Large Language Models

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Talking About Large Language Models — mentioned by corford

Mon, 09 Jan 2023 11:36:05 GMT

I have tried it and it's great. But, because it's so effective, it's easy to over extrapolate and anthropomorphise it. I think it's useful to keep in perspective (as the article and referenced paper - https://arxiv.org/abs/2212.03551 - talk about) that ChatGPT/LLMs are, ultimately, "just" extremely powerful statistical next-token prediction systems. That doesn't at all invalidate the achievement or belittle the impact such systems will have on society but, keeping this in mind does help to avoid blowing everything out of proportion (which is easy to do because, yep, ChatGPT is genuinely very…

Paper and discussion history. A mention is not an endorsement.

Talking About Large Language Models — mentioned by LewisVerstappen

Fri, 06 Jan 2023 01:40:56 GMT

Oh yeah sure. Actually, a research scientist at DeepMind wrote a paper around this - clarifying some of the misconceptions around LLMs (people implying anthromorphisms that don't really exit). https://arxiv.org/pdf/2212.03551.pdf Guess it's just prevented by educating people on what these models are doing under the hood?

Paper and discussion history. A mention is not an endorsement.

Talking About Large Language Models — mentioned by neonate

Sat, 10 Dec 2022 19:49:41 GMT

https://arxiv.org/pdf/2212.03551.pdf

Paper and discussion history. A mention is not an endorsement.

Talking About Large Language Models — HN submission

Sat, 10 Dec 2022 16:12:02 GMT

Talking About Large Language Models

146 points; 149 discussion replies on HN. Paper and source history. A submission is not an endorsement.