paper-with-me

Papers

Analyzing Attention Mechanisms through Lens of Sample Complexity and Loss Landscape

2021-01-01 · Bingyuan Liu, Yogesh Balaji, Lingzhou Xue, Martin Renqiang Min

Attention mechanisms have advanced state-of-the-art deep learning models in many machine learning tasks. Despite significant empirical gains, there is a lack of theoretical analyses on their effectiveness. In this paper, we address this problem by studying the sample complexity and loss landscape of attention-based neural networks. Our results show that, under mild assumptions, every local minimum of the attention model has low prediction error, and attention models require lower sample complexity than models without attention. Besides revealing why popular self-attention works, our theoretical results also provide guidelines for designing future attention models. Experiments on various datasets validate our theoretical findings.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

BIG-bench Machine Learning

Similar Papers 제목 키워드 기반

A Framework of Defining, Modeling, and Analyzing Cognition Mechanisms

2023-11-13 · Amir Fayezioghani

Cognition is a core part of and a common topic among philosophy of mind, psychology, neuroscience, AI, and cognitive science. Through a mechanistic lens, I propose a framework of defining, modeling, and analyzing cogniti…

Philosophy

Attention Lens: A Tool for Mechanistically Interpreting the Attention Head Information Retrieval Mechanism

2023-10-25 · Mansi Sakarvadia, Arham Khan, Aswathy Ajith, Daniel Grzenda 외

Transformer-based Large Language Models (LLMs) are the state-of-the-art for natural language tasks. Recent work has attempted to decode, by reverse engineering the role of linear layers, the internal mechanisms by which …

Information RetrievalRetrieval

Analyzing Memorization in Large Language Models through the Lens of Model Attribution

2025-01-09 · Tarun Ram Menta, Susmit Agrawal, Chirag Agarwal

Large Language Models (LLMs) are prevalent in modern applications but often memorize training data, leading to privacy breaches and copyright issues. Existing research has mainly focused on posthoc analyses, such as extr…

Memorization

Positional LSH: Binary Block Matrix Approximation for Attention with Linear Biases

2026-05-10 · Daniel Wolfson, Tal Wagner arxiv

Positional encoding in transformers is commonly implemented through positional embeddings, attention masks, or bias terms, but formal connections between these mechanisms remain limited. We study attention with positiona…

Probing the Probing Paradigm: Does Probing Accuracy Entail Task Relevance?

2020-05-02 · EACL 2021 2 · Abhilasha Ravichander, Yonatan Belinkov, Eduard Hovy

Although neural models have achieved impressive results on several NLP benchmarks, little is understood about the mechanisms they use to perform language tasks. Thus, much recent attention has been devoted to analyzing t…

Natural Language InferenceSentenceWord Embeddings