Meta-in-context learning in large language models — HN submission

Wed, 24 May 2023 12:38:51 GMT

Google researchers: non-editable LLMs can be modified by chatting, no finetuning

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

Meta-in-context learning in large language models — mentioned by verdverm

Tue, 23 May 2023 21:23:32 GMT

https://arxiv.org/abs/2305.12907 > We find that meta-in-context learning adaptively modifies priors over latent variables, ultimately leading to priors that closely resemble the true statistics of the environment. Furthermore, our analysis reveals that meta-in-context learning can not only be used to change prior expectations but is also capable of reshaping an LLM’s learning strategies This was done using OpenAI's public facing APIs Even if it is not a permanent "edit" it still influences the model at the lowest levels

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

Meta-in-context learning in large language models · HackerPapers discussions

Meta-in-context learning in large language models — HN submission

Meta-in-context learning in large language models — mentioned by verdverm