mlsanityData to 14 Oct 2026

GLEAN: Active Generalized Category Discovery with Diverse LLM Feedback

arXiv 2502.18414 ↗First posted 25 Feb 2025cs.CL cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

Generalized Category Discovery (GCD) is a practical and challenging open-world task that aims to recognize both known and novel categories in unlabeled data using limited labeled data from known categories. Due to the lack of supervision, previous GCD methods face significant challenges, such as difficulty in rectifying errors for confusing instances, and inability to effectively uncover and leverage the semantic meanings of discovered clusters. Therefore, additional annotations are usually required for real-world applicability. However, human annotation is extremely costly and inefficient. To address these issues, we propose GLEAN, a unified framework for generalized category discovery that actively learns from diverse and collaborative LLM feedback. Our approach leverages three different types of LLM feedback to: (1) improve instance-level contrastive features, (2) generate category descriptions, and (3) align uncertain instances with LLM-selected category descriptions. Extensive experiments demonstrate the superior performance of GLEAN over state-of-the-art models across diverse datasets, metrics, and supervision settings. Our code is available at https://github.com/amazon-science/Glean.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 27 of its 44 references that are in the map. First posted on arXiv 25 Feb 2025.

Builds on

27 of 44 references.

References · 27

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2015Unsupervised Deep Embedding for Clustering Analysiscited by 767
  3. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  4. 2020A Survey of Deep Active Learningcited by 554
  5. 2020GoEmotions: A Dataset of Fine-Grained Emotionscited by 486
  6. 2020Efficient Intent Detection with Dual Sentence Encoderscited by 433
  7. 2019An Evaluation Dataset for Intent Classification and Out-of-Scope Predictioncited by 386
  8. 2018FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluationcited by 254
  9. 2020Contextual Diversity for Active Learningcited by 139
  10. 2021Neighborhood Contrastive Learning for Novel Class Discoverycited by 121
  11. 2022Parametric Classification for Generalized Category Discovery: A Baseline Studycited by 113
  12. 2020OpenMix: Reviving Known Knowledge for Discovering Novel Visual Categories in An Open Worldcited by 93
15 more
  1. 2023ClusterLLM: Large Language Models as a Guide for Text Clusteringcited by 70
  2. 2020Discovering New Intents with Deep Aligned Clusteringcited by 68
  3. 2019Discovering New Intents via Constrained Deep Adaptive Clustering with Cluster Refinementcited by 56
  4. 2022Generalized Category Discovery with Decoupled Prototypical Networkcited by 35
  5. 2023Improving Contrastive Learning of Sentence Embeddings from AI Feedbackcited by 29
  6. 2022New Intent Discovery with Pre-training and Contrastive Learningcited by 28
  7. 2021OntoED: Low-resource Event Detection with Ontology Embeddingcited by 23
  8. 2023IDAS: Intent Discovery with Abstractive Summarizationcited by 22
  9. 2024ImplicitAVE: An Open-Source Dataset and Multimodal LLMs Benchmark for Implicit Attribute Value Extractioncited by 17
  10. 2023JointMatch: A Unified Approach for Diverse and Collaborative Pseudo-Labeling to Semi-Supervised Text Classificationcited by 15
  11. 2023Generalized Category Discovery with Large Language Models in the Loopcited by 15
  12. 2024EIVEN: Efficient Implicit Attribute Value Extraction using Multimodal LLMcited by 12
  13. 2023DeCrisisMB: Debiased Semi-Supervised Learning for Crisis Tweet Classification via Memory Bankcited by 10
  14. 2023Pre-training Intent-Aware Encoders for Zero- and Few-Shot Intent Classificationcited by 5
  15. 2023RSVP: Customer Intent Detection via Agent Response Contrastive and Generative Pre-Trainingcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 5; 2026: 1.

6 papers.

  1. 2025Scaling Laws for Many-Shot In-Context Learning with Self-Generated Annotationscited by 7
  2. 2025TestNUC: Enhancing Test-Time Computing Approaches and Scaling through Neighboring Unlabeled Data Consistencycited by 4
  3. 2025LLMs Enable Bag-of-Texts Representations for Short-Text Clusteringcited by 2
  4. 2025Generalized Category Discovery in Event-Centric Contexts: Latent Pattern Mining with LLMscited by 0
  5. 2025Lifelong Evolution: Collaborative Learning between Large and Small Language Models for Continuous Emergent Fake News Detectioncited by 0
  6. 2026OmniGCD: Abstracting Generalized Category Discovery for Modality Agnosticismcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Learning Like Humans: Analogical Concept Learning for Generalized Category Discovery
  2. 2026CloSeR: Unified Relational Distillation from Closed-Set Teachers for Category Discovery
  3. 2025Prior-Constrained Association Learning for Fine-Grained Generalized Category Discovery
  4. 2025HIDISC: A Hyperbolic Framework for Domain Generalization with Generalized Category Discovery
  5. 2025DebGCD: Debiased Learning with Distribution Guidance for Generalized Category Discovery
  6. 2026Generalized Category Discovery under Domain Shifts: From Vision to Vision-Language Models
  7. 2026Test-Time Generalized Category Discovery
  8. 2025ReLKD: Inter-Class Relation Learning with Knowledge Distillation for Generalized Category Discovery