Elamur Research⌕
☰
home / papers / 2020 / Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning

Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning

Paper · 2020 · Lianhui Qin, Vered Shwartz, Peter West et al.

Summary

Abductive reasoning asks what plausibly happened between an observed past and an observed future. Counterfactual reasoning asks how a story's later events should change when an earlier event is altered. Both require a model to condition on context on both sides of the text it generates. Standard left-to-right language models such as GPT2 only see the preceding context, and text-infilling models handle only short spans.

The authors propose DELOREAN, a decoding method for logical reasoning of the nonmonotonic kind (its name is an acronym formed from those words). It requires no supervision and works with an off-the-shelf left-to-right LM:

  • A forward pass produces output logits from the past context.
  • A backward pass back-propagates a task-specific loss tied to the future constraint. It adjusts only the output representation, not the model weights.
  • The two sets of logits are blended, and the cycle repeats.
  • Candidates from the iterations are ranked for coherence with BERT's next sentence prediction.

The method is evaluated on the ART dataset (abductive NLG) and the TIME TRAVEL dataset (counterfactual story rewriting). DELOREAN beats all unsupervised baselines in automatic and human evaluation, and its outputs are judged more coherent than those of some supervised systems. This suggests pretrained LMs can perform nonmonotonic reasoning without task-specific training data.

Key points

  • How it works: DELOREAN alternates forward passes, which roughly favour P(Y|X), with backward passes, which roughly push toward P(Z|XY). Parameters of the pretrained LM (GPT2) stay frozen throughout.
  • Benchmarks: ART (about 20k observation pairs taken from ROCStories) and TIME TRAVEL (five-sentence stories with a counterfactual change to the second sentence).
  • Counterfactual results: DELOREAN balances coherence and minimal edits. It is more coherent than discriminative baselines (Recon+CF, Sup+Disc) and needs fewer edits than zero-shot re-ranked baselines. Recon+CF copied the original ending verbatim in 84% of cases.
  • Human evaluation: inter-rater agreement was fair, with Fleiss' κ = 0.34.
  • Code: released at github.com/qkaren/unsup_gen_for_cms_reasoning.

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

🎓University of WashingtonSeattle · United States263 works · 180,581 citations🔬Allen Institute for Artificial IntelligenceSeattle · United States219 works · 89,712 citations

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 9,086 worksArtificial Intelligence · 8,889 worksArts and Humanities · 9,296 worksLanguage and Linguistics · 9,206 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 510 works

Cite

Cite as Lianhui Qin, Vered Shwartz, Peter West, Chandra Bhagavatula, Jena D. Hwang, Ronan Le Bras, Antoine Bosselut and Yejin Choi. 2020. Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2020.emnlp-main.58

@inproceedings{qin2020back,
  title = {{Back to the Future: Unsupervised Backprop-based Decoding for Counterfactual and Abductive Commonsense Reasoning}},
  author = {Lianhui Qin, Vered Shwartz, Peter West, Chandra Bhagavatula, Jena D. Hwang, Ronan {Le Bras}, Antoine Bosselut and Yejin Choi},
  booktitle = {Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)},
  year = {2020},
  doi = {10.18653/v1/2020.emnlp-main.58},
  url = {https://doi.org/10.18653/v1/2020.emnlp-main.58}
}

References

57 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Jessica Ficler and Yoav Goldberg. 2017. Controlling linguistic style aspects in neural language generation. In Proceedings of the Workshop on Stylistic Variation, pages 94–104.
  2. Henning Andersen. 1973. Abductive and deductive change. Language, pages 765–793. find
  3. Mikel Artetxe, Gorka Labaka, Eneko Agirre, and Kyunghyun Cho. 2017. Unsupervised neural machine translation. arXiv preprint arXiv:1710.11041.
  4. Yu Bao, Hao Zhou, Shujian Huang, Lei Li, Lili Mou, Olga Vechtomova, Xinyu Dai, and Jiajun Chen. 2019. Generating sentences from disentangled syntactic and semantic spaces. arXiv preprint arXiv:1907.05789.
  5. Christos Baziotis, Ion Androutsopoulos, Ioannis Konstas, and Alexandros Potamianos. 2019. Seq3: Differentiable sequence-to-sequence-to-sequence autoencoder for unsupervised abstractive sentence compression. In NAACL-HLT. find
  6. Chandra Bhagavatula, Ronan Le Bras, Chaitanya Malaviya, Keisuke Sakaguchi, Ari Holtzman, Hannah Rashkin, Doug Downey, Wen-tau Yih, and Yejin Choi. 2019. Abductive commonsense reasoning. In International Conference on Learning Representations.
  7. Angela Fan, Mike Lewis, and Yann Dauphin. 2018. Hierarchical neural story generation. In ACL. find
  8. Joseph L Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological bulletin, 76(5):378. find
  9. Matthew L Ginsberg. 1986. Counterfactuals. Artificial intelligence, 30(1):35–79. find
  10. Nelson Goodman. 1947. The problem of counterfactual conditionals. The Journal of Philosophy, 44(5):113–128. find
  11. Mark Granroth-Wilding and Stephen Clark. 2016. What happens next? event prediction using a compositional neural network model. In Thirtieth AAAI Conference on Artificial Intelligence. find
  12. Jerry R Hobbs, Mark E Stickel, Douglas E Appelt, and Paul Martin. 1993. Interpretation as abduction. Artificial intelligence, 63(1-2):69–142. find
  13. Zhiting Hu, Zichao Yang, Xiaodan Liang, Ruslan Salakhutdinov, and Eric P Xing. 2017. Toward controlled generation of text. In ICML.
  14. Antoine Bosselut, Hannah Rashkin, Maarten Sap, Chaitanya Malaviya, Asli Celikyilmaz, and Yejin Choi. 2019. COMET: Commonsense transformers for automatic knowledge graph construction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4762–4779, Florence…
  15. Lifu Huang, Ronan Le Bras, Chandra Bhagavatula, and Yejin Choi. 2019. Cosmos qa: Machine reading comprehension with contextual commonsense reasoning. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Langua…
  16. Samuel R Bowman, Gabor Angeli, Christopher Potts, and Christopher D Manning. 2015. A large annotated corpus for learning natural language inference. arXiv preprint arXiv:1508.05326. find
  17. Steve D Isard. 1974. What would you have done if...? Theoretical Linguistics, 1(1-3):233–256. find
  18. Nathanael Chambers and Dan Jurafsky. 2008. Unsupervised learning of narrative event chains. In Proceedings of ACL-08: HLT, pages 789–797. find
  19. Sumanth Dathathri, Andrea Madotto, Janice Lan, Jane Hung, Eric Frank, Piero Molino, Jason Yosinski, and Rosanne Liu. 2019. Plug and play language models: a simple approach to controlled text generation. arXiv preprint arXiv:1912.02164.
  20. Philip Nicholas Johnson-Laird. 2006. How we reason. Oxford University Press, USA. find
  21. Nitish Shirish Keskar, Bryan McCann, Lav R Varshney, Caiming Xiong, and Richard Socher. 2019. Ctrl: A conditional transformer language model for controllable generation. arXiv preprint arXiv:1909.05858.
  22. Rik Koncel-Kedziorski, Ioannis Konstas, Luke Zettlemoyer, and Hannaneh Hajishirzi. 2016. A theme-rewriting approach for generating algebra word problems. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1617–1628. find
  23. Sarit Kraus, Daniel Lehmann, and Menachem Magidor. 1990. Nonmonotonic reasoning, preferential models and cumulative logics. Artificial intelligence, 44(1-2):167–207. find
  24. Guillaume Lample, Myle Ott, Alexis Conneau, Ludovic Denoyer, and Marc’Aurelio Ranzato. 2018. Phrase-based & neural unsupervised machine translation. arXiv preprint arXiv:1804.07755. find
  25. Guillaume Lample, Sandeep Subramanian, Eric Smith, Ludovic Denoyer, Marc’Aurelio Ranzato, and Y-Lan Boureau. 2019. Multiple-attribute text rewriting. In ICLR. find
  26. Carolin Lawrence and Stefan Riezler. 2018. Improving a neural semantic parser by counterfactual learning from human bandit feedback. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1820–1830. find
  27. Carolin Lawrence, Artem Sokolov, and Stefan Riezler. 2017. Counterfactual learning from bandit feedback under deterministic logging: A case study in statistical machine translation. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 2566–2576. find
  28. Zhongyang Li, Xiao Ding, and Ting Liu. 2018. Constructing narrative event evolutionary graph for script event prediction. In IJCAI. find
  29. Chin-Yew Lin. 2004. Rouge: A package for automatic evaluation of summaries. Text Summarization Branches Out.
  30. Hugo Mercier and Dan Sperber. 2017. The enigma of reason. Harvard University Press. find
  31. Ning Miao, Hao Zhou, Lili Mou, Rui Yan, and Lei Li. 2019. Cgmh: Constrained sentence generation by metropolis-hastings sampling. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, pages 6834–6842. find
  32. Nasrin Mostafazadeh, Nathanael Chambers, Xiaodong He, Devi Parikh, Dhruv Batra, Lucy Vanderwende, Pushmeet Kohli, and James Allen. 2016a. A corpus and evaluation framework for deeper understanding of commonsense stories. arXiv preprint arXiv:1604.01696. find
  33. Nasrin Mostafazadeh, Nathanael Chambers, Xiaodong He, Devi Parikh, Dhruv Batra, Lucy Vanderwende, Pushmeet Kohli, and James F. Allen. 2016b. A corpus and cloze evaluation for deeper understanding of commonsense stories. In HLT-NAACL. find
  34. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU: a method for automatic evaluation of machine translation. In ACL, pages 311– 318.
  35. Judea Pearl and Dana Mackenzie. 2018. The book of why: the new science of cause and effect. Basic Books. find
  36. Charles Sanders Peirce. 1960. Collected papers of charles sanders peirce, volume 2. Harvard University Press. find
  37. Karl Pichotta and Raymond Mooney. 2014. Statistical script learning with multi-argument events. In EACL, pages 220–229. find
  38. Lianhui Qin, Antoine Bosselut, Ari Holtzman, Chandra Bhagavatula, Elizabeth Clark, and Yejin Choi. 2019a. Counterfactual story reasoning and generation. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Lan… find
  39. Lianhui Qin, Michel Galley, Chris Brockett, Xiaodong Liu, Xiang Gao, Bill Dolan, Yejin Choi, and Jianfeng Gao. 2019b. Conversing by reading: Contentful neural conversation with on-demand machine reading. In ACL, pages 5427–5436. find
  40. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019a. Language models are unsupervised multitask learners. -. find
  41. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019b. Language models are unsupervised multitask learners. OpenAI Blog, 1:8. find
  42. Hannah Rashkin, Antoine Bosselut, Maarten Sap, Kevin Knight, and Yejin Choi. 2018. Modeling naive psychology of characters in simple commonsense stories. arXiv preprint arXiv:1805.06533. find
  43. Raymond Reiter. 1988. Nonmonotonic reasoning. In Exploring artificial intelligence, pages 439–481. Elsevier. find
  44. Melissa Roemmele, Cosmin Adrian Bejan, and Andrew S Gordon. 2011. Choice of plausible alternatives: An evaluation of commonsense causal reasoning. In 2011 AAAI Spring Symposium Series. find
  45. Maarten Sap, Ronan LeBras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A Smith, and Yejin Choi. 2019. ATOMIC: An atlas of machine commonsense for if-then reasoning. In AAAI. find
  46. Roger C Schank and Robert P Abelson. 1977. Scripts, plans, goals and understanding: An inquiry into human knowledge structures. find
  47. Raphael Schumann, Lili Mou, Yao Lu, Olga Vechtomova, and Katja Markert. 2020. Discrete optimization for unsupervised sentence summarization with word-level extraction. arXiv preprint arXiv:2005.01791. find
  48. Tianxiao Shen, Tao Lei, Regina Barzilay, and Tommi Jaakkola. 2017. Style transfer from non-parallel text by cross-alignment. In Advances in neural information processing systems, pages 6830–6841.
  49. Youngseo Son, Anneke Buffone, Joe Raso, Allegra Larche, Anthony Janocko, Kevin Zembroski, H Andrew Schwartz, and Lyle Ungar. 2017. Recognizing counterfactual thinking in social media texts. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Pa… find
  50. William Starr. 2019. Counterfactuals. In Edward N. Zalta, editor, The Stanford Encyclopedia of Philosophy, fall 2019 edition. Metaphysics Research Lab, Stanford University. find
  51. Yoel Zeldes, Dan Padnos, and Barak Peleg. 2020. Haim-1.5 - the next generation. find
  52. Qing Sun, Stefan Lee, and Dhruv Batra. 2017. Bidirectional beam search: Forward-backward inference in neural sequence models for fill-in-the-blank image captioning. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 6961–6969. find
  53. Niket Tandon, Bhavana Dalvi Mishra, Keisuke Sakaguchi, Antoine Bosselut, and Peter Clark. 2019. Wiqa: A dataset for” what if...” reasoning over procedural text. In EMNLP. find
  54. Peter West, Ari Holtzman, Jan Buys, and Yejin Choi. find
  55. Rowan Zellers, Ari Holtzman, Hannah Rashkin, Yonatan Bisk, Ali Farhadi, Franziska Roesner, and Yejin Choi. 2019. Defending against neural fake news. In Advances in Neural Information Processing Systems, pages 9051–9062.
  56. Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. 2019. BERTScore: Evaluating text generation with BERT. CoRR, abs/1904.09675. find
  57. Wanrong Zhu, Zhiting Hu, and Eric Xing. 2019. Text infilling. arXiv preprint arXiv:1901.00158. — [arXiv:1901.00158](https://arxiv.org/abs/1901.00158) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/qin-2020-back-future-unsupervised-backprop-based/