Many-Shot In-Context Learning · HackerPapers discussions

Many-Shot In-Context Learning — mentioned by blixt

Tue, 15 Jul 2025 06:56:34 GMT

This is one type of problem of information retrieval, but I think the change in performance with context length may be different for non-retrieval answers (such as “what is the edited code for making this button red?” or “which of the above categories does the sentence ‘…’ fall under?”). One paper that stood out to me a while back was Many-Shot In-Context Learning[1] which showed large positive jumps in performance from filling the context with examples. As always, it’s important to test one’s problem to know how the LLM changes in behavior for different context contents/lengths — I wouldn’t…

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

Many-Shot In-Context Learning — mentioned by nl

Tue, 30 Apr 2024 23:53:11 GMT

> context length doesn't unlock intelligence, just more information This isn't correct (for most definitions of "unlock intelligence"). In-Context Learning (ICL) can do everything off-line training can do. > We find that both Reinforced and Unsupervised ICL can be quite effective in the many-shot regime, particularly on complex reasoning tasks. Finally, we demonstrate that, unlike few-shot learning, many-shot learning is effective at overriding pretraining biases and can learn high-dimensional functions with numerical inputs. https://arxiv.org/abs/2404.11018

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

Many-Shot In-Context Learning — HN submission

Mon, 22 Apr 2024 00:38:15 GMT

Many-Shot In-Context Learning

62 points; 1 discussion reply on HN. Paper and source history. A submission is not an endorsement.