Lu, Sheng ; Bigoulaeva, Irina ; Sachdeva, Rachneet ; Madabushi, Harish Tayyar ; Gurevych, Iryna (2024)
Are Emergent Abilities in Large Language Models just In-Context Learning?
62nd Annual Meeting of the Association for Computational Linguistics. Bangkok, Thailand (11.08.2024 - 16.08.2024)
doi: 10.18653/v1/2024.acl-long.279
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

Large language models, comprising billions of parameters and pre-trained on extensive web-scale corpora, have been claimed to acquire certain capabilities without having been specifically trained on them. These capabilities, referred to as “emergent abilities,” have been a driving force in discussions regarding the potentials and risks of language models. A key challenge in evaluating emergent abilities is that they are confounded by model competencies that arise through alternative prompting techniques, including in-context learning, which is the ability of models to complete a task based on a few examples. We present a novel theory that explains emergent abilities, taking into account their potential confounding factors, and rigorously substantiate this theory through over 1000 experiments. Our findings suggest that purported emergent abilities are not truly emergent, but result from a combination of in-context learning, model memory, and linguistic knowledge. Our work is a foundational step in explaining language model performance, providing a template for their efficient use and clarifying the paradox of their ability to excel in some instances while faltering in others. Thus, we demonstrate that their capabilities should not be overestimated.

Frage zum Eintrag

Redaktionelle Details anzeigen