mlsanityData to 7 Oct 2026

Contrastive Instruction Tuning

arXiv 2402.11138 ↗First posted 17 Feb 2024cs.CL cs.AI cs.LGcited by 17

Not on the map. Why?

Abstract (arXiv)

Instruction tuning has been used as a promising approach to improve the performance of large language models (LLMs) on unseen tasks. However, current LLMs exhibit limited robustness to unseen instructions, generating inconsistent outputs when the same instruction is phrased with slightly varied forms or language styles. This behavior indicates LLMs' lack of robustness to textual variations and generalizability to unseen instructions, potentially leading to trustworthiness issues. Accordingly, we propose Contrastive Instruction Tuning, which maximizes the similarity between the hidden representations of semantically equivalent instruction-instance pairs while minimizing the similarity between semantically different ones. To facilitate this approach, we augment the existing FLAN collection by paraphrasing task instructions. Experiments on the PromptBench benchmark show that CoIN consistently improves LLMs' robustness to unseen instructions with variations across character, word, sentence, and semantic levels by an average of +2.5% in accuracy. Code is available at https://github.com/luka-group/CoIN.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 35 of its 60 references that are in the map. First posted on arXiv 17 Feb 2024.

Builds on

35 of 60 references.

References · 35

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,854
  5. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
  6. 2017A Broad-Coverage Challenge Corpus for Sentence Understanding through Inferencecited by 2,665
  7. 2015Character-level Convolutional Networks for Text Classificationcited by 2,293
  8. 2021SimCSE: Simple Contrastive Learning of Sentence Embeddingscited by 2,168
  9. 2019BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questionscited by 2,140
  10. 2018Know What You Don't Know: Unanswerable Questions for SQuADcited by 1,538
  11. 2017SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluationcited by 998
  12. 2020Beyond Accuracy: Behavioral Testing of NLP models with CheckListcited by 979
23 more
  1. 2018Neural Network Acceptability Judgmentscited by 952
  2. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
  3. 2019Adversarial NLI: A New Benchmark for Natural Language Understandingcited by 798
  4. 2023The Flan Collection: Designing Data and Methods for Effective Instruction Tuningcited by 734
  5. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 696
  6. 2019Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailmentcited by 663
  7. 2023Instruction Tuning for Large Language Models: A Surveycited by 617
  8. 2023How Far Can Camels Go? Exploring the State of Instruction Tuning on Open Resourcescited by 405
  9. 2019PAWS: Paraphrase Adversaries from Word Scramblingcited by 395
  10. 2018WiC: the Word-in-Context Dataset for Evaluating Context-Sensitive Meaning Representationscited by 370
  11. 2018Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifierscited by 341
  12. 2021Exploring the Limits of Out-of-Distribution Detectioncited by 283
  13. 2021Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Modelscited by 227
  14. 2023PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Promptscited by 218
  15. 2020Contrastive Training for Improved Out-of-Distribution Detectioncited by 186
  16. 2022BRIO: Bringing Order to Abstractive Summarizationcited by 161
  17. 2022Controllable Natural Language Generation with Contrastive Prefixescited by 75
  18. 2023Evaluating the Zero-shot Robustness of Instruction-tuned Language Modelscited by 63
  19. 2022ZeroPrompt: Scaling Prompt-Based Pretraining to 1,000 Tasks Improves Zero-Shot Generalizationcited by 54
  20. 2023Active Instruction Tuning: Improving Cross-Task Generalization by Training on Prompt Sensitive Taskscited by 53
  21. 2022Robustness of Learning from Task Instructionscited by 34
  22. 2023Exploring Format Consistency for Instruction Tuningcited by 14
  23. 2023Learning to Generalize for Cross-domain QAcited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 11; 2026: 5.

17 papers.

Sort:

By area:

  1. 2025reWordBench: Benchmarking and Improving the Robustness of Reward Models with Transformed Inputscited by 25
  2. 2025RobustFlow: Towards Robust Agentic Workflow Generationcited by 19
  3. 2025Refine Knowledge of Large Language Models via Adaptive Contrastive Learningcited by 14
  4. 2025MCIF: Multimodal Crosslingual Instruction-Following Benchmark from Scientific Talkscited by 14
  5. 2025RaCoT: Plug-and-Play Contrastive Example Generation Mechanism for Enhanced LLM Reasoning Reliabilitycited by 14
  6. 2025Order Matters: Investigate the Position Bias in Multi-constraint Instruction Followingcited by 12
  7. 2025Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metricscited by 12
  8. 2024Monotonic Paraphrasing Improves Generalization of Language Model Promptingcited by 10
  9. 2026Towards Self-Robust LLMs: Intrinsic Prompt Noise Resistance via CoIPOcited by 5
  10. 2025Flip-Flop Consistency: Unsupervised Training for Robustness to Prompt Perturbations in LLMscited by 3
  11. 2025Fine-Tuning on Noisy Instructions: Effects on Generalization and Performancecited by 2
  12. 2026Prompting Underestimates LLM Capability for Time Series Classificationcited by 1
5 more
  1. 2025CL3DOR: Contrastive Learning for 3D Large Multimodal Models via Odds Ratio on High-Resolution Point Cloudscited by 0
  2. 2025Waste Not, Want Not; Recycled Gumbel Noise Improves Consistency in Natural Language Generationcited by 0
  3. 2026SCENIC: Semantic-Conditioned Edge-Aware Neural Framework for Structured IoT Command Generationcited by 0
  4. 2026Cross-Relational Preference Learning for Better LLM Instruction Followingcited by 0
  5. 2026Prompt-Robust Language Models: Which Training Strategies Work?cited by 0