The False Promise of Imitating Proprietary LLMs · HackerPapers discussions

The False Promise of Imitating Proprietary LLMs — HN submission

Thu, 25 Jun 2026 03:56:59 GMT

The False Promise of Imitating Proprietary LLMs

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

The False Promise of Imitating Proprietary LLMs — mentioned by nonethewiser

Thu, 25 Jun 2026 01:21:52 GMT

Distilled models are necessarily behind so long as models are progressing. Models are progressing. Maybe it will be over some time in the future. And Berkeley’s “False Promise of Imitating Proprietary LLMs” found imitation closes the style gap fast but there is a large capability gap. https://arxiv.org/abs/2305.15717

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

The False Promise of Imitating Proprietary LLMs — mentioned by adt

Wed, 21 Jun 2023 05:32:25 GMT

Exactly. The False Promise of Imitating Proprietary LLMs (UC Berkeley, 25/May/2023) https://arxiv.org/abs/2305.15717

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

The False Promise of Imitating Proprietary LLMs — mentioned by yyyk

Mon, 05 Jun 2023 10:11:53 GMT

>I think we don't need to argue the fact that an on premise solution is cheaper than a cloud solution for a lot of tasks In the absolute sense where we look at the total cost of running the model (and not care how it's distributed or include profits), you may be right even with scale efficiencies - making a determination requires data about cloud server costs we do not have. But I can make an informed guess about the dollar cost the user sees. The cost the user sees is influenced by the factor called 'Microsoft and Google (etc.) have a lot of money, and seem to be perfectly willing to absorb…

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

The False Promise of Imitating Proprietary LLMs — mentioned by finexplained

Wed, 31 May 2023 11:24:47 GMT

tldr: LoRA fine-tuning of an open-source alpaca model. See https://arxiv.org/abs/2305.15717 for why calling this a private ChatGPT is probably misleading.

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

The False Promise of Imitating Proprietary LLMs — mentioned by ma2rten

Sat, 27 May 2023 01:48:44 GMT

https://arxiv.org/abs/2305.15717

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

The False Promise of Imitating Proprietary LLMs — mentioned by throwaway55905

Fri, 26 May 2023 23:19:15 GMT

Perhaps it's not a moat. However, if the advantage is due to things like inference infrastructure to support a massive model, that isn't easy to duplicate. I would also say that the quality of these smaller models are good, but we also may not be measuring them correctly. Recent papers suggest that these smaller LMs dont fully capture ChatGPT quality in ways that may not have appeared with crowd worker ratings [1]. It's easy to have your inputs be inside a happy distribution for a paper but fail in the real world in ways that GPT-4 doesnt. Lmsys would love to compare with bigger models but…

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

The False Promise of Imitating Proprietary LLMs — mentioned by elexhobby

Fri, 26 May 2023 19:34:38 GMT

For another possibility, see https://arxiv.org/abs/2305.15717. The new models may not actually be better - the evaluation may be broken.

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

The False Promise of Imitating Proprietary LLMs — HN submission

Fri, 26 May 2023 01:42:45 GMT

The False Promise of Imitating Proprietary LLMs

126 points; 83 discussion replies on HN. Paper and source history. A submission is not an endorsement.