paper-with-me

홈 › Papers

The representation landscape of few-shot learning and fine-tuning in large language models

2024-09-05 · Diego Doimo, Alessandro Serra, Alessio Ansuini, Alberto Cazzaniga

In-context learning (ICL) and supervised fine-tuning (SFT) are two common strategies for improving the performance of modern large language models (LLMs) on specific tasks. Despite their different natures, these strategies often lead to comparable performance gains. However, little is known about whether they induce similar representations inside LLMs. We approach this problem by analyzing the probability landscape of their hidden representations in the two cases. More specifically, we compare how LLMs solve the same question-answering task, finding that ICL and SFT create very different internal structures, in both cases undergoing a sharp transition in the middle of the network. In the first half of the network, ICL shapes interpretable representations hierarchically organized according to their semantic content. In contrast, the probability landscape obtained with SFT is fuzzier and semantically mixed. In the second half of the model, the fine-tuned representations develop probability modes that better encode the identity of answers, while the landscape of ICL representations is characterized by less defined peaks. Our approach reveals the diverse computational strategies developed inside LLMs to solve the same task across different conditions, allowing us to make a step towards designing optimal methods to extract information from language models.

📄 PDF Abstract BibTeX arXiv:2409.03662

Code (1)

diegodoimo/geometry_icl_finetuning 공식 구현 pytorch

Tasks

Few-Shot LearningIn-Context LearningQuestion Answering

Methods 이 논문이 사용한 방법론

SFT Shrink and Fine-Tune, or SFT, is a type of distillation that avoids explicit distillation by copying parameters to a student student model and then fine-tuning.…

Similar Papers 제목 키워드 기반

Flatten Long-Range Loss Landscapes for Cross-Domain Few-Shot Learning

2024-03-01 · CVPR 2024 1 · Yixiong Zou, Yicong Liu, Yiman Hu, Yuhua Li 외

Cross-domain few-shot learning (CDFSL) aims to acquire knowledge from limited training data in the target domain by leveraging prior knowledge transferred from source domains with abundant training samples. CDFSL faces c…

Cross-Domain Few-Shotcross-domain few-shot learningFew-Shot Learning

Navigating the Landscape of Large Language Models: A Comprehensive Review and Analysis of Paradigms and Fine-Tuning Strategies

2024-04-13 · Benjue Weng

With the surge of ChatGPT,the use of large models has significantly increased,rapidly rising to prominence across the industry and sweeping across the internet. This article is a comprehensive review of fine-tuning metho…

Few-Shot LearningKnowledge DistillationMulti-Task Learningparameter-efficient fine-tuning

Transfer Learning for Finetuning Large Language Models

2024-11-02 · Tobias Strangmann, Lennart Purucker, Jörg K. H. Franke, Ivo Rapant 외

As the landscape of large language models expands, efficiently finetuning for specific tasks becomes increasingly crucial. At the same time, the landscape of parameter-efficient finetuning methods rapidly expands. Conseq…

Bayesian OptimizationMeta-LearningTransfer Learning

Exploring The Landscape of Distributional Robustness for Question Answering Models

2022-10-22 · Anas Awadalla, Mitchell Wortsman, Gabriel Ilharco, Sewon Min 외

We conduct a large empirical evaluation to investigate the landscape of distributional robustness in question answering. Our investigation spans over 350 models and 16 question answering datasets, including a diverse set…

In-Context LearningQuestion Answering

Delta Activations: A Representation for Finetuned Large Language Models

2025-09-04 · Zhiqiu Xu, Amish Sethi, Mayur Naik, Ser-Nam Lim arxiv

The success of powerful open source Large Language Models (LLMs) has enabled the community to create a vast collection of post-trained models adapted to specific tasks and domains. However, navigating and understanding t…