Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Out-of-Distribution Detection by Leveraging Between-Layer Transformation Smoothness (2310.02832v2)

Published 4 Oct 2023 in cs.LG and cs.CL

Abstract: Effective out-of-distribution (OOD) detection is crucial for reliable machine learning models, yet most current methods are limited in practical use due to requirements like access to training data or intervention in training. We present a novel method for detecting OOD data in Transformers based on transformation smoothness between intermediate layers of a network (BLOOD), which is applicable to pre-trained models without access to training data. BLOOD utilizes the tendency of between-layer representation transformations of in-distribution (ID) data to be smoother than the corresponding transformations of OOD data, a property that we also demonstrate empirically. We evaluate BLOOD on several text classification tasks with Transformer networks and demonstrate that it outperforms methods with comparable resource requirements. Our analysis also suggests that when learning simpler tasks, OOD data transformations maintain their original sharpness, whereas sharpness increases with more complex tasks.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (60)
  1. State-of-the-art in artificial neural network applications: A survey. Heliyon, 4(11), 2018.
  2. Classical out-of-distribution detection methods benchmark in text classification tasks. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop), pp. 119–129, Toronto, Canada, July 2023. Association for Computational Linguistics. doi: 10.18653/v1/2023.acl-srw.20. URL https://aclanthology.org/2023.acl-srw.20.
  3. Norm and trace estimation with random rank-one vectors. SIAM Journal on Matrix Analysis and Applications, 42(1):202–223, 2021.
  4. One billion word benchmark for measuring progress in statistical language modeling. pp.  2635 – 2639, 2014. doi: 10.21437/interspeech.2014-564.
  5. ELECTRA: Pre-training text encoders as discriminators rather than generators. In International Conference on Learning Representations, 2020. URL https://openreview.net/forum?id=r1xMH1BtvB.
  6. Aleatory or epistemic? Does it matter? Structural safety, 31(2):105–112, 2009.
  7. An image is worth 16x16 words: Transformers for image recognition at scale. In 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7, 2021. OpenReview.net, 2021. URL https://openreview.net/forum?id=YicbFdNTTy.
  8. Tom Fawcett. An introduction to roc analysis. Pattern recognition letters, 27(8):861–874, 2006.
  9. Yarin Gal. Uncertainty in deep learning. 2016.
  10. Dropout as a bayesian approximation: Representing model uncertainty in deep learning. In International conference on machine learning, pp. 1050–1059. PMLR, 2016.
  11. What do vision transformers learn? a visual exploration. arXiv preprint arXiv:2212.06727, 2022.
  12. Igeood: An information geometry approach to out-of-distribution detection. In International Conference on Learning Representations, 2022.
  13. On calibration of modern neural networks. In International conference on machine learning, pp. 1321–1330. PMLR, 2017.
  14. A baseline for detecting misclassified and out-of-distribution examples in neural networks. In ICLR, 2017.
  15. Deep anomaly detection with outlier exposure. In ICLR, 2019.
  16. Generalized odin: Detecting out-of-distribution image without learning from out-of-distribution data. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp.  10951–10960, 2020.
  17. On the importance of gradients for detecting distributional shifts in the wild. Advances in Neural Information Processing Systems, 34:677–689, 2021.
  18. M.F. Hutchinson. A stochastic estimator of the trace of the influence matrix for laplacian smoothing splines. Communications in Statistics - Simulation and Computation, 19(2):433–450, 1990.
  19. What does BERT learn about the structure of language? In ACL 2019-57th Annual Meeting of the Association for Computational Linguistics, 2019.
  20. What uncertainties do we need in bayesian deep learning for computer vision? Advances in neural information processing systems, 30, 2017.
  21. Adam: A method for stochastic optimization. In Yoshua Bengio and Yann LeCun (eds.), 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015. URL http://arxiv.org/abs/1412.6980.
  22. On pitfalls in ood detection: Entropy considered harmful, 2021. Uncertainty & Robustness in Deep Learning Workshop, ICML.
  23. Back to the basics: Revisiting out-of-distribution detection baselines. arXiv preprint arXiv:2207.03061, 2022.
  24. Simple and scalable predictive uncertainty estimation using deep ensembles. Advances in neural information processing systems, 30, 2017.
  25. Ken Lang. Newsweeder: Learning to filter netnews. In Machine learning proceedings 1995, pp.  331–339. Elsevier, 1995.
  26. Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11):2278–2324, 1998.
  27. A simple unified framework for detecting out-of-distribution samples and adversarial attacks. Advances in neural information processing systems, 31, 2018.
  28. Datasets: A community library for natural language processing. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pp.  175–184, Online and Punta Cana, Dominican Republic, November 2021. Association for Computational Linguistics. doi: 10.18653/v1/2021.emnlp-demo.21. URL https://aclanthology.org/2021.emnlp-demo.21.
  29. Learning question classifiers. In COLING 2002: The 19th International Conference on Computational Linguistics, 2002. URL https://www.aclweb.org/anthology/C02-1150.
  30. Enhancing the reliability of out-of-distribution image detection in neural networks. In 6th International Conference on Learning Representations, ICLR 2018, 2018.
  31. Simple and principled uncertainty estimation with deterministic deep learning via distance awareness. Advances in Neural Information Processing Systems, 33:7498–7512, 2020a.
  32. Energy-based out-of-distribution detection. Advances in neural information processing systems, 33:21464–21475, 2020b.
  33. RoBERTa: A robustly optimized BERT pretraining approach. arXiv preprint arXiv:1907.11692, 2019.
  34. Learning word vectors for sentiment analysis. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, pp.  142–150, Portland, Oregon, USA, June 2011. Association for Computational Linguistics. URL http://www.aclweb.org/anthology/P11-1015.
  35. Image-based recommendations on styles and substitutes. In Proceedings of the 38th international ACM SIGIR conference on research and development in information retrieval, pp.  43–52, 2015.
  36. A common language effect size statistic. Psychological bulletin, 111(2):361, 1992.
  37. Classification uncertainty of deep neural networks based on gradient information. In Artificial Neural Networks in Pattern Recognition: 8th IAPR TC3 Workshop, ANNPR 2018, Siena, Italy, September 19–21, 2018, Proceedings 8, pp.  113–125. Springer, 2018.
  38. Can you trust your model’s uncertainty? Evaluating predictive uncertainty under dataset shift. Advances in neural information processing systems, 32, 2019.
  39. Bo Pang and Lillian Lee. A sentimental education: Sentiment analysis using subjectivity summarization based on minimum cuts. In Proceedings of the ACL, 2004.
  40. Pytorch: An imperative style, high-performance deep learning library. In H. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alché-Buc, E. Fox, and R. Garnett (eds.), Advances in Neural Information Processing Systems, volume 32. Curran Associates, Inc., 2019. URL https://proceedings.neurips.cc/paper_files/paper/2019/file/bdbca288fee7f92f2bfa9f7012727740-Paper.pdf.
  41. Scikit-learn: Machine learning in python. the Journal of machine Learning research, 12:2825–2830, 2011.
  42. Rissanen data analysis: Examining dataset characteristics via description length. In International Conference on Machine Learning, pp. 8500–8513. PMLR, 2021.
  43. Dissecting contextual word embeddings: Architecture and representation. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pp.  1499–1509, 2018.
  44. Revisiting mahalanobis distance for transformer-based out-of-domain detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pp.  13675–13682, 2021.
  45. A case for new neural network smoothness constraints. In Jessica Zosa Forde, Francisco Ruiz, Melanie F. Pradier, and Aaron Schein (eds.), Proceedings on "I Can’t Believe It’s Not Better!" at NeurIPS Workshops, volume 137 of Proceedings of Machine Learning Research, pp.  21–32. PMLR, 12 Dec 2020. URL https://proceedings.mlr.press/v137/rosca20a.html.
  46. Does your dermatology classifier know what it doesn’t know? detecting the long-tail of unseen conditions. Medical Image Analysis, 75:102274, 2022.
  47. BIGPATENT: A large-scale dataset for abstractive and coherent summarization. CoRR, abs/1906.03741, 2019. URL http://arxiv.org/abs/1906.03741.
  48. Recursive deep models for semantic compositionality over a sentiment treebank. In Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pp.  1631–1642, Seattle, Washington, USA, October 2013. Association for Computational Linguistics. URL https://www.aclweb.org/anthology/D13-1170.
  49. Dropout: a simple way to prevent neural networks from overfitting. The journal of machine learning research, 15(1):1929–1958, 2014.
  50. Dice: Leveraging sparsification for out-of-distribution detection. In European Conference on Computer Vision, pp.  691–708. Springer, 2022.
  51. React: Out-of-distribution detection with rectified activations. Advances in Neural Information Processing Systems, 34:144–157, 2021.
  52. Out-of-distribution detection with deep nearest neighbors. In International Conference on Machine Learning, pp. 20827–20840. PMLR, 2022.
  53. Dataset cartography: Mapping and diagnosing datasets with training dynamics. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp.  9275–9293, 2020.
  54. BERT rediscovers the classical nlp pipeline. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pp.  4593–4601, 2019.
  55. An effective baseline for robustness to distributional shift. In 2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA), pp.  278–285. IEEE, 2021.
  56. Uncertainty estimation using a single deep deterministic neural network. In International conference on machine learning, pp. 9690–9700. PMLR, 2020.
  57. Attention is all you need. Advances in neural information processing systems, 30, 2017.
  58. Transformers: State-of-the-art natural language processing. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pp.  38–45, Online, October 2020. Association for Computational Linguistics. doi: 10.18653/v1/2020.emnlp-demos.6. URL https://aclanthology.org/2020.emnlp-demos.6.
  59. Generalized out-of-distribution detection: A survey. arXiv preprint arXiv:2110.11334, 2021.
  60. Character-level convolutional networks for text classification. Advances in neural information processing systems, 28, 2015.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (5)
  1. Fran Jelenić (3 papers)
  2. Josip Jukić (10 papers)
  3. Martin Tutek (10 papers)
  4. Mate Puljiz (18 papers)
  5. Jan Šnajder (24 papers)
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now