Lune

EMNLP Findings2023

Retrieval-Augmented Few-shot Text Classification

Guoxin Yu, Lemao Liu, Haiyun Jiang, Shuming Shi, Xiang Ao

Abstract

Retrieval-augmented methods are successful in the standard scenario where the retrieval space is sufficient; whereas in the few-shot scenario with limited retrieval space, this paper shows it is non-trivial to put them into practice. First, it is impossible to retrieve semantically similar examples by using an off-the-shelf metric and it is crucial to learn a task-specific retrieval metric; Second, our preliminary experiments demonstrate that it is difficult to optimize a plausible metric by minimizing the standard cross-entropy loss. The in-depth analyses quantitatively show minimizing cross-entropy loss suffers from the weak supervision signals and the severe gradient vanishing issue during the optimization. To address these issues, we introduce two novel training objectives, namely EM-L and R-L, which provide more task-specific guidance to the retrieval metric by the EM algorithm and a ranking-based loss, respectively. Extensive experiments on 10 datasets prove the superiority of the proposed retrieval augmented methods on the performance.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

5 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Improving Language Models by Retrieving from Trillions of Tokens cites Language Models are Few-Shot Learners
  • Improving Language Models by Retrieving from Trillions of Tokens cites Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
  • Knowledge Boundary Discovery for Large Language Models cites Language Models are Few-Shot Learners
  • FeVisQA: Free-Form Question Answering over Data Visualizations cites Towards Robustness of Text-to-Visualization Translation Against Lexical and Phrasal Variability
  • FeVisQA: Free-Form Question Answering over Data Visualizations cites Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
  • Language Models are Few-Shot Learners cites UnifiedQA: Crossing Format Boundaries With a Single QA System
  • Language Models are Few-Shot Learners cites Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
  • Neuron-Aware Active Few-Shot Learning for LLMs cites Language Models are Few-Shot Learners
  • AI Chatbots as Professional Service Agents: Developing a Professional Identity cites Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers5

  1. Towards Robustness of Text-to-Visualization Translation Against Lexical and Phrasal VariabilityJinwei Lu, Yuanfeng Song, Haodi Zhang, Chen Jason Zhang et al.ICDE 2025 · 3 citations
  2. FeVisQA: Free-Form Question Answering over Data VisualizationsYuanfeng Song, Jinwei Lu, Yuanwei Song, Caleb Chen Cao et al.ICDE 2025 · 2 citations
  3. Neuron-Aware Active Few-Shot Learning for LLMsZhuowei Chen, Liwei Chen, Christian Schunn, Raquel Coelho et al.ACL 2026 · 1 citation
  4. AI Chatbots as Professional Service Agents: Developing a Professional IdentityWenwen Li, Kangwei Shi, Yidong ChaiEMNLP 2025 · 1 citation
  5. Knowledge Boundary Discovery for Large Language ModelsZiquan Wang, Zhongqi LuAAAI 2026

Builds on24

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. Retrieval-Augmented Generation for Knowledge-Intensive NLP TasksPatrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni et al.NeurIPS 2020 · 19,162 citations
  3. Improving Language Models by Retrieving from Trillions of TokensSebastian Borgeaud, Arthur Mensch, Jordan Hoffmann, Trevor Cai et al.ICML 2022 · 1,629 citations
  4. UnifiedQA: Crossing Format Boundaries With a Single QA SystemDaniel Khashabi, Sewon Min, Tushar Khot, Ashish Sabharwal et al.EMNLP Findings 2020 · 514 citations
  5. Template-Based Named Entity Recognition Using BARTLeyang Cui, Yu Wu, Jian Liu, Sen Yang et al.ACL Findings 2021 · 342 citations

Related papers

  1. Boosting Few-Shot Learning With Adaptive Margin LossAoxue Li, Weiran Huang, Xu Lan, Jiashi Feng et al.CVPR 2020
  2. Few-Shot Recognition via Stage-Wise Retrieval-Augmented FinetuningTian Liu, Huixin Zhang, Shubham Parashar, Shu KongCVPR 2025
  3. SEC-Prompt: SEmantic Complementary Prompting for Few-Shot Class-Incremental LearningYe Liu, Meng YangCVPR 2025
  4. Semantic Prompt for Few-Shot Image RecognitionCVPR 2023
  5. Semantic-based Selection, Synthesis, and Supervision for Few-shot LearningJinda Lu, Shuo Wang, Xinyu Zhang, Yanbin Hao et al.ACM MM 2023 · 7 citations
Dusk over the sea between two cliffs drawn in fine vertical lines