Larger language models do in-context learning differently

Jerry Wei, Jason Wei, Yi Tay, Dustin Tran, Albert Webson, Yifeng Lu, Xinyun Chen, Hanxiao Liu, Da Huang, Denny Zhou, Tengyu Ma

Primary Area: generative models

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: In-context learning, natural language processing, large language models

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: We study how in-context learning in language models is affected by semantic priors versus input-label mappings.

Abstract: We study how in-context learning (ICL) in language models is affected by semantic priors versus input-label mappings. We investigate two setups - ICL with flipped labels and ICL with semantically-unrelated labels - across various model families (GPT-3, InstructGPT, Codex, an internal model, and an instruction-tuned variant of the internal model). First, experiments on ICL with flipped labels show that overriding semantic priors is an emergent ability of model scale. While small language models ignore flipped labels presented in-context and thus rely primarily on semantic priors from pretraining, large models can override semantic priors when presented with in-context exemplars that contradict priors, despite the stronger semantic priors that larger models may hold. We next study semantically-unrelated label ICL (SUL-ICL), in which labels are semantically unrelated to their inputs (e.g., foo/bar instead of negative/positive), thereby forcing language models to learn the input-label mappings shown in in-context exemplars in order to perform the task. The ability to do SUL-ICL also emerges primarily with scale, and large-enough language models can even perform linear classification in a SUL-ICL setting. Finally, we evaluate instruction-tuned models and find that instruction tuning strengthens both the use of semantic priors and the capacity to learn input-label mappings, but more of the former.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 4092