paper-with-me

홈 › Papers

Are Emergent Abilities in Large Language Models just In-Context Learning?

2023-09-04 · Sheng Lu, Irina Bigoulaeva, Rachneet Sachdeva, Harish Tayyar Madabushi, Iryna Gurevych

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.

📄 PDF Abstract BibTeX arXiv:2309.01809

Code (1)

ukplab/on-emergence 공식 구현 pytorch

Tasks

In-Context LearningInstruction FollowingLanguage ModelingLanguage Modelling

Similar Papers 제목 키워드 기반

A Theory of Emergent In-Context Learning as Implicit Structure Induction

2023-03-14 · Michael Hahn, Navin Goyal

Scaling large language models (LLMs) leads to an emergent capacity to learn in-context from example demonstrations. Despite progress, theoretical understanding of this phenomenon remains limited. We argue that in-context…

In-Context Learning

Emergent Abilities in Large Language Models: A Survey

2025-02-28 · Leonardo Berti, Flavio Giorgi, Gjergji Kasneci

Large Language Models (LLMs) are leading a new technological revolution as one of the most promising research streams toward artificial general intelligence. The scaling of these models, accomplished by increasing the nu…

In-Context LearningMisconceptionsSurvey

Emergent Abilities of Large Language Models

2022-06-15 · Jason Wei, Yi Tay, Rishi Bommasani, Colin Raffel 외

Scaling up language models has been shown to predictably improve performance and sample efficiency on a wide range of downstream tasks. This paper instead discusses an unpredictable phenomenon that we refer to as emergen…

Language Modelling

Emergent Abilities in Reduced-Scale Generative Language Models

2024-04-02 · Sherin Muckatira, Vijeta Deshpande, Vladislav Lialin, Anna Rumshisky

Large language models can solve new tasks without task-specific fine-tuning. This ability, also known as in-context learning (ICL), is considered an emergent ability and is primarily seen in large language models with bi…

In-Context LearningZero-Shot Learning

Training on the Test Task Confounds Evaluation and Emergence

2024-07-10 · Ricardo Dominguez-Olmedo, Florian E. Dorner, Moritz Hardt

We study a fundamental problem in the evaluation of large language models that we call training on the test task. Unlike wrongful practices like training on the test data, leakage, or data contamination, training on the …

BenchmarkingLanguage Modelling