Elamur Research⌕
☰
home / papers / 2026 / Active Generalized Category Discovery with Diverse LLM Feedback

Active Generalized Category Discovery with Diverse LLM Feedback

Paper · 2026 · Henry Peng Zou, Siffi Singh, Yi Nian et al.

Summary

The paper addresses Generalized Category Discovery (GCD) for text. In GCD, a model must sort unlabeled data into both previously known and entirely new categories, using only a small labeled set drawn from the known ones. Earlier GCD approaches combine contrastive learning with K-Means++ clustering. Without extra supervision, they cannot fix mistakes on confusing examples or attach meaning to the clusters they find, and human annotation is too expensive.

The authors propose DeLFGCD, which uses an LLM as a cheap annotator for the most uncertain instances, selected by clustering entropy. The LLM provides three kinds of feedback:

  • it picks the most similar neighbor for each ambiguous example, which feeds neighborhood contrastive learning;
  • it writes names and descriptions for discovered clusters;
  • it assigns ambiguous examples to one of those descriptions, so that instance embeddings can be aligned with category text.

At inference time, only the fine-tuned encoder and K-Means++ are used. On the BANKING, CLINC and StackOverflow intent datasets, the method reports new state-of-the-art results across supervision settings. It costs under $1 per run, which makes interpretable open-world categorization practical.

Key points

  • Cost: With gpt-4o-mini and a default budget of 500 queried ambiguous samples, one run uses roughly 1.55M–5.88M tokens. This costs about $0.25–$0.91, and all LLM querying finishes in under 15 minutes.
  • Query budget: Adding more queried samples helps. On CLINC, ARI rises from 77.5% to 85.0% as queries grow from 0 to 2000, while gains on BANKING level off around 500.
  • Comparison with GPT-4o: DeLFGCD beats GPT-4o used directly, even when GPT-4o is given all category names (KCR=100%), with or without in-context demonstrations. DeLFGCD reaches 76.98/80.26 accuracy on BANKING (KCR 25%/50%) versus 70.67–71.67 for GPT-4o. On CLINC it reaches 91.51/94.53 versus 80.67–86.00.
  • Class imbalance: On imbalanced BANKING (KCR 25%), LLM feedback improves novel classes more than known ones (+26.07% vs +14.42%). It also improves both head and tail classes.
  • Robustness and clustering: Results are stable over three runs, with an accuracy standard deviation of about 0.64 on BANKING. Replacing K-Means++ with GMM or HDBSCAN gives similar performance. The framework is currently limited to text.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🏢AmazonSeattle · United States112 works · 40,565 citations🎓University of Illinois ChicagoChicago · United States61 works · 20,107 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,320 works

Cite

Cite as Henry Peng Zou, Siffi Singh, Yi Nian, Jianfeng He, Jason Cai, Saab Mansour and Hang Su. 2026. Active Generalized Category Discovery with Diverse LLM Feedback. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-long.358

@inproceedings{zou2026active,
  title = {{Active Generalized Category Discovery with Diverse LLM Feedback}},
  author = {Henry Peng Zou, Siffi Singh, Yi Nian, Jianfeng He, Jason Cai, Saab Mansour and Hang Su},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-long.358},
  url = {https://doi.org/10.18653/v1/2026.eacl-long.358}
}

References

46 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sharat Agarwal, Himanshu Arora, Saket Anand, and Chetan Arora. 2020. Contextual diversity for active learning. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVI 16, pages 137–153. Springer. find
  2. Wenbin An, Wenkai Shi, Feng Tian, Haonan Lin, QianYing Wang, Yaqiang Wu, Mingxiang Cai, Luyan Wang, Yan Chen, Haiping Zhu, and Ping Chen. 2024. Generalized category discovery with large language models in the loop. In Findings of the Association for Computational Linguistics ACL 2024, pages 8653–866… find
  3. Wenbin An, Feng Tian, Qinghua Zheng, Wei Ding, QianYing Wang, and Ping Chen. 2023. Generalized category discovery with decoupled prototypical network. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 12527– 12535. find
  4. Jordan T. Ash, Chicheng Zhang, Akshay Krishnamurthy, John Langford, and Alekh Agarwal. 2020. Deep batch active learning by diverse, uncertain gradient lower bounds. In International Conference on Learning Representations. find
  5. Jianhong Bai, Zuozhu Liu, Hualiang Wang, Ruizhe Chen, Lianrui Mu, Xiaomeng Li, Joey Tianyi Zhou, YANG FENG, Jian Wu, and Haoji Hu. 2023. Towards distribution-agnostic generalized category discovery. In Thirty-seventh Conference on Neural Information Processing Systems. find
  6. Iñigo Casanueva, Tadas Temčinas, Daniela Gerz, Matthew Henderson, and Ivan Vulić. 2020. Efficient intent detection with dual sentence encoders. In Proceedings of the 2nd Workshop on Natural Language Processing for Conversational AI, pages 38–45, Online. Association for Computational Linguistics.
  7. Qinyuan Cheng, Xiaogui Yang, Tianxiang Sun, Linyang Li, and Xipeng Qiu. 2023. Improving contrastive learning of sentence embeddings from AI feedback. In Findings of the Association for Computational Linguistics: ACL 2023, pages 11122–11138, Toronto, Canada. Association for Computational Linguistics. find
  8. Dorottya Demszky, Dana Movshovitz-Attias, Jeongwoo Ko, Alan Cowen, Gaurav Nemade, and Sujith Ravi. 2020. Goemotions: A dataset of fine-grained emotions. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4040–4054. find
  9. Shumin Deng, Ningyu Zhang, Luoqiu Li, Chen Hui, Tou Huaixiao, Mosha Chen, Fei Huang, and Huajun Chen. 2021. Ontoed: Low-resource event detection with ontology embedding. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confer… find
  10. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In North American Chapter of the Association for Computational Linguistics.
  11. Shansan Gong, Zelin Zhou, Shuo Wang, Fengjiao Chen, Xiujie Song, Xuezhi Cao, Yunsen Xian, and Kenny Zhu. 2023. Transferable and efficient: Unifying dynamic multi-domain product categorization. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Indus… find
  12. Xu Han, Hao Zhu, Pengfei Yu, Ziyun Wang, Yuan Yao, Zhiyuan Liu, and Maosong Sun. 2018. Fewrel: A large-scale supervised few-shot relation classification dataset with state-of-the-art evaluation. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 4803–480… find
  13. Sheng-Jun Huang, Rong Jin, and Zhi-Hua Zhou. 2010. Active learning by querying informative and representative examples. Advances in neural information processing systems, 23. find
  14. Harold W Kuhn. 1955. The hungarian method for the assignment problem. Naval research logistics quarterly, 2(1-2):83–97. find
  15. Stefan Larson, Anish Mahendran, Joseph J. Peper, Christopher Clarke, Andrew Lee, Parker Hill, Jonathan K. Kummerfeld, Kevin Leach, Michael A. Laurenzano, Lingjia Tang, and Jason Mars. 2019. An evaluation dataset for intent classification and out-of-scope prediction. In Proceedings of the 2019 Confer… find
  16. Jinggui Liang and Lizi Liao. 2023. ClusterPrompt: Cluster semantic enhanced prompt learning for new intent discovery. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 10468–10481, Singapore. Association for Computational Linguistics. find
  17. Jinggui Liang, Lizi Liao, Hao Fei, and Jing Jiang. 2024a. Synergizing large language models and pre-trained smaller models for conversational intent discovery. In Findings of the Association for Computational Linguistics: ACL 2024, pages 14133–14147, Bangkok, Thailand. Association for Computational… find
  18. Jinggui Liang, Lizi Liao, Hao Fei, Bobo Li, and Jing Jiang. 2024b. Actively learn from LLMs with uncertainty propagation for generalized category discovery. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologi… find
  19. Ting-En Lin, Hua Xu, and Hanlei Zhang. 2019. Discovering new intents via constrained deep adaptive clustering with cluster refinement. ArXiv, abs/1911.08891. — [arXiv:1911.08891](https://arxiv.org/abs/1911.08891) find
  20. Dan Wang and Yi Shang. 2014. A new active labeling method for deep learning. In 2014 International joint conference on neural networks (IJCNN), pages 112– 119. IEEE. find
  21. Shijie Ma, Fei Zhu, Zhun Zhong, Xu-Yao Zhang, and Cheng-Lin Liu. 2024a. Active generalized category discovery. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16890–16900. find
  22. Xin Wen, Bingchen Zhao, and Xiaojuan Qi. 2023. Parametric classification for generalized category discovery: A baseline study. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 16590–16600. find
  23. Shijie Ma, Fei Zhu, Zhun Zhong, Xu-Yao Zhang, and Cheng-Lin Liu. 2024b. Active generalized category discovery. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 16890– 16900. find
  24. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, Mariama Drame, Quentin Lho… — Transformers: State-of-the-Art Natural Language Processing
  25. Nan Pu, Zhun Zhong, and Nicu Sebe. 2023. Dynamic conceptional contrastive learning for generalized category discovery. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 7579–7588. find
  26. Maarten De Raedt, Fréderic Godin, Thomas Demeester, and Chris Develder. 2023. Idas: Intent discovery with abstractive summarization. ArXiv, abs/2305.19783. find
  27. Hannah Rashkin, Eric Michael Smith, Margaret Li, and Y-Lan Boureau. 2019. Towards empathetic open-domain conversation models: A new benchmark and dataset. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 5370–5381. find
  28. Pengzhen Ren, Yun Xiao, Xiaojun Chang, Po-Yao Huang, Zhihui Li, Brij B Gupta, Xiaojiang Chen, and Xin Wang. 2021. A survey of deep active learning. ACM computing surveys (CSUR), 54(9):1–40. find
  29. Ozan Sener and Silvio Savarese. 2017. Active learning for convolutional neural networks: A core-set approach. arXiv preprint arXiv:1708.00489. find
  30. Mujeen Sung, James Gung, Elman Mansimov, Nikolaos Pappas, Raphael Shu, Salvatore Romeo, Yi Zhang, and Vittorio Castelli. 2023. Pre-training intent-aware encoders for zero- and few-shot intent classification. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, p… find
  31. Yu-Chien Tang, Wei-Yao Wang, An-Zi Yen, and Wen-Chih Peng. 2023. RSVP: Customer intent detection via agent response contrastive and generative pre-training. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 10400– 10412, Singapore. Association for Computational Linguist… find
  32. Sagar Vaze, Kai Han, Andrea Vedaldi, and Andrew Zisserman. 2022. Generalized category discovery. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7492– 7501. find
  33. Junyuan Xie, Ross Girshick, and Ali Farhadi. 2016. Unsupervised deep embedding for clustering analysis. In International conference on machine learning, pages 478–487. PMLR. find
  34. Jiaming Xu, Peng Wang, Guanhua Tian, Bo Xu, Jun Zhao, Fangyuan Wang, and Hongwei Hao. 2015. Short text clustering via convolutional neural networks. In Proceedings of the 1st Workshop on Vector Space Modeling for Natural Language Processing, pages 62–69, Denver, Colorado. Association for Computation… find
  35. Feng Zhang, Wei Chen, Fei Ding, Meng Gao, Tengjiao Wang, Jiahui Yao, and Jiabin Zheng. 2024. From discrimination to generation: Low-resource intent detection with language model instruction tuning. In Findings of the Association for Computational Linguistics ACL 2024, pages 10167–10183, Bangkok, Tha… find
  36. Hanlei Zhang, Hua Xu, Ting-En Lin, and Rui Lyu. 2021. Discovering new intents with deep aligned clustering. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 14365–14373. find
  37. Hanlei Zhang, Hua Xu, Xin Wang, Fei Long, and Kai Gao. 2023a. A clustering framework for unsupervised and semi-supervised new intent discovery. IEEE Transactions on Knowledge and Data Engineering. find
  38. Yuwei Zhang, Zihan Wang, and Jingbo Shang. 2023b. ClusterLLM: Large language models as a guide for text clustering. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 13903–13920, Singapore. Association for Computational Linguistics. find
  39. Yuwei Zhang, Haode Zhang, Li-Ming Zhan, Xiao-Ming Wu, and Albert Lam. 2022. New intent discovery with pre-training and contrastive learning. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 256–269, Dublin, Ireland. Association… find
  40. Zhun Zhong, Enrico Fini, Subhankar Roy, Zhiming Luo, Elisa Ricci, and Nicu Sebe. 2021a. Neighborhood contrastive learning for novel class discovery. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10867–10875. find
  41. Zhun Zhong, Linchao Zhu, Zhiming Luo, Shaozi Li, Yi Yang, and Nicu Sebe. 2021b. Openmix: Reviving known knowledge for discovering novel visual categories in an open world. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9462–9470. find
  42. Yunhua Zhou, Guofeng Quan, and Xipeng Qiu. 2023. A probabilistic framework for discovering new intents. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 3771–3784, Toronto, Canada. Association for Computational Linguistics. find
  43. Henry Zou and Cornelia Caragea. 2023. JointMatch: A unified approach for diverse and collaborative pseudo-labeling to semi-supervised text classification. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7290–7301, Singapore. Association for Computatio… find
  44. Henry Zou, Vinay Samuel, Yue Zhou, Weizhi Zhang, Liancheng Fang, Zihe Song, Philip Yu, and Cornelia Caragea. 2024a. ImplicitAVE: An open-source dataset and multimodal LLMs benchmark for implicit attribute value extraction. In Findings of the Association for Computational Linguistics ACL 2024, pages… find
  45. Henry Zou, Gavin Yu, Ziwei Fan, Dan Bu, Han Liu, Peng Dai, Dongmei Jia, and Cornelia Caragea. 2024b. EIVEN: Efficient implicit attribute value extraction using multimodal LLM. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human… find
  46. Henry Zou, Yue Zhou, Weizhi Zhang, and Cornelia Caragea. 2023. DeCrisisMB: Debiased semi-supervised learning for crisis tweet classification via memory bank. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 6104–6115, Singapore. Association for Computational Linguistic… find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zou-2026-active-generalized-category-discovery-diverse/