mlsanityData to 14 Oct 2026

TestNUC: Enhancing Test-Time Computing Approaches and Scaling through Neighboring Unlabeled Data Consistency

arXiv 2502.19163 ↗First posted 26 Feb 2025cs.CL cs.AI cs.IRcited by 4

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

Test-time computing approaches, which leverage additional computational resources during inference, have been proven effective in enhancing large language model performance. This work introduces a novel, linearly scaling approach, TestNUC, that improves test-time predictions by leveraging the local consistency of neighboring unlabeled data-it classifies an input instance by considering not only the model's prediction on that instance but also on neighboring unlabeled instances. We evaluate TestNUC across eight diverse datasets, spanning intent classification, topic mining, domain discovery, and emotion detection, demonstrating its consistent superiority over baseline methods such as standard prompting and self-consistency. Furthermore, TestNUC can be seamlessly integrated with existing test-time computing approaches, substantially boosting their performance. Our analysis reveals that TestNUC scales effectively with increasing amounts of unlabeled data and performs robustly across different embedding models, making it practical for real-world applications. Our code is available at https://github.com/HenryPengZou/TestNUC.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 30 of its 43 references that are in the map. First posted on arXiv 26 Feb 2025.

Builds on

30 of 43 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug

Other references · 29

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  3. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  4. 2023Let's Verify Step by Stepcited by 3,789
  5. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,085
  6. 2019MixMatch: A Holistic Approach to Semi-Supervised Learningcited by 1,391
  7. 2020Making Pre-trained Language Models Better Few-shot Learnerscited by 1,256
  8. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 974
  9. 2019Generalization through Memorization: Nearest Neighbor Language Modelscited by 800
  10. 2020GoEmotions: A Dataset of Fine-Grained Emotionscited by 486
  11. 2024NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Modelscited by 472
  12. 2020Efficient Intent Detection with Dual Sentence Encoderscited by 433
17 more
  1. 2019An Evaluation Dataset for Intent Classification and Out-of-Scope Predictioncited by 386
  2. 2020SCAN: Learning to Classify Images without Labelscited by 352
  3. 2023Label Words are Anchors: An Information Flow Perspective for Understanding In-Context Learningcited by 236
  4. 2020MTOP: A Comprehensive Multilingual Task-Oriented Semantic Parsing Benchmarkcited by 158
  5. 2023Compositional Exemplars for In-context Learningcited by 144
  6. 2023Few-shot Fine-tuning vs. In-context Learning: A Fair Comparison and Evaluationcited by 136
  7. 2023ClusterLLM: Large Language Models as a Guide for Text Clusteringcited by 70
  8. 2024Revisiting Demonstration Selection Strategies in In-Context Learningcited by 58
  9. 2019Meta-Learning with Dynamic-Memory-Based Prototypical Network for Few-Shot Event Detectioncited by 54
  10. 2023kNN Prompting: Beyond-Context Learning with Calibration-Free Nearest Neighbor Inferencecited by 47
  11. 2023Self-ICL: Zero-Shot In-Context Learning with Self-Generated Demonstrationscited by 42
  12. 2024A Survey on LLM Inference-Time Self-Improvementcited by 17
  13. 2024On the Noise Robustness of In-Context Learning for Text Generationcited by 16
  14. 2023JointMatch: A Unified Approach for Diverse and Collaborative Pseudo-Labeling to Semi-Supervised Text Classificationcited by 15
  15. 2023DeCrisisMB: Debiased Semi-Supervised Learning for Crisis Tweet Classification via Memory Bankcited by 10
  16. 2025GLEAN: Active Generalized Category Discovery with Diverse LLM Feedbackcited by 6
  17. 2024BANGS: Game-Theoretic Node Selection for Graph Self-Trainingcited by 2

Built on by

Cited by year

Citing papers per year: 2025: 4.

4 papers.

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 384
  2. 2025From Web Search towards Agentic Deep Research: Incentivizing Search with Reasoning Agentscited by 51
  3. 2025Scaling Laws for Many-Shot In-Context Learning with Self-Generated Annotationscited by 7
  4. 2025Diagnosing Failures in Large Language Models' Answers: Integrating Error Attribution into Evaluation Frameworkcited by 3

Similar, not cited

Close in wording, no citation either way

  1. 2026Test-Time Augmentation for LLMs: When Input Diversity Beats Output Diversity at Matched Compute
  2. 2025First Finish Search: Efficient Test-Time Scaling in Large Language Models
  3. 2025A Survey on Test-Time Scaling in Large Language Models: What, How, Where, and How Well?
  4. 2025Representation Consistency for Accurate and Coherent LLM Answer Aggregation
  5. 2025The Art of Scaling Test-Time Compute for Large Language Models
  6. 2026SCATR: Simple Calibrated Test-Time Ranking
  7. 2025Understanding the Role of Training Data in Test-Time Scaling
  8. 2025Sample Complexity and Representation Ability of Test-time Scaling Paradigms