Classification or Generation? Understanding Paradigm Shift for Knowledge-Intensive Tasks
Knowledge-intensive tasks such as entity retrieval are challenging for even cutting edge NLP models since they require models to apply knowledge about the world. Previous studies typically treat this task as classification. Recently, a new paradigm has emerged, which reformats knowledge-intensive tasks as natural language generation. This post summarizes the paradigm shift and reviews the new generative methodology for the ICLR community, providing philosophical questions and new directions.
Code (0)
등록된 구현이 없습니다.
Tasks
ClassificationEntity RetrievalRetrievalText GenerationSimilar Papers 제목 키워드 기반
Active Perception Agent for Omnimodal Audio-Video Understanding
Omnimodal large language models have made significant strides in unifying audio and visual modalities; however, they often face challenges in fine-grained cross-modal understanding and have difficulty with multimodal ali…
Response GenerationCLoG: Benchmarking Continual Learning of Image Generation Models
Continual Learning (CL) poses a significant challenge in Artificial Intelligence, aiming to mirror the human ability to incrementally acquire knowledge and skills. While extensive research has focused on CL within the co…
BenchmarkingContinual LearningDiversityImage Generation+1Algorithmic Recourse in the Wild: Understanding the Impact of Data and Model Shifts
As predictive models are increasingly being deployed to make a variety of consequential decisions, there is a growing emphasis on designing algorithms that can provide recourse to affected individuals. Existing recourse …
Computational Intelligence Challenges and Applications on Large-Scale Astronomical Time Series Databases
Time-domain astronomy (TDA) is facing a paradigm shift caused by the exponential growth of the sample size, data complexity and data generation rates of new astronomical sky surveys. For example, the Large Synoptic Surve…
AstronomyBIG-bench Machine LearningTime SeriesTime Series AnalysisMeta-causal Learning for Single Domain Generalization
Single domain generalization aims to learn a model from a single training domain (source domain) and apply it to multiple unseen test domains (target domains). Existing methods focus on expanding the distribution of the …
counterfactualCounterfactual InferenceDomain Generalizationimage-classification+3