Elamur Research⌕
☰
home / papers / 2023 / ReCEval: Evaluating Reasoning Chains via Correctness and Informativeness

ReCEval: Evaluating Reasoning Chains via Correctness and Informativeness

Paper · 2023 · Archiki Prasad, Swarnadeep Saha, Xiang Zhou et al.

Summary

Language models increasingly produce step-by-step rationales, but they are usually judged only by whether the final answer is right. That can hide flawed or shortcut reasoning. Reference-based checks against human-written chains are also costly, and gold chains are rarely unique. The authors propose ReCEval, a reference-free framework that treats a reasoning chain as an informal proof and scores it on two properties:

  • Correctness: whether each step's conclusion follows from its own premises (intra-step) and stays consistent with the input context and earlier steps (inter-step).
  • Informativeness: whether each step adds new information that moves the chain toward the answer.

Steps are split into Reasoning Content Units (RCUs), which are claim-level pieces separated into premises and a conclusion. Correctness is scored with NLI models, and informativeness with pointwise V-information (PVI) computed using T5-large.

In meta-evaluation against perturbed and human-annotated errors on Entailment Bank, GSM-8K and DROP, ReCEval generally correlates with errors better than prior metrics such as ROSCOE. Using ReCEval scores to choose among sampled chains also raises task accuracy. This makes it a practical tool for diagnosing and improving reasoning quality.

Key points

  • On Entailment Bank, ReCEval achieves the top correlation for 5 of 6 error types. For hallucinations, correlation rises from 0.62 to 0.89.
  • On the overall quality measure, correlation improves from 0.28 to 0.36 on GSM-8K and from 0.19 to 0.22 on DROP. Across these two datasets, ReCEval leads on 5 of 7 error types.
  • On GSM-8K with FLAN T5-XXL, the authors sample 20 chains and pick the best-ranked one:
    • Greedy decoding: 17.3% accuracy.
    • Correctness only: 19.6%.
    • Informativeness only: 18.7%.
    • Both properties combined: 20.5% (+3.2 points).
    • ROSCOE combinations gain at most 1.7 points.
  • The framework does not depend on a particular model. Prompting GPT-3.5-turbo for scores worked on a 50-chain Entailment Bank subset.
  • Stated limitations: steps that rely heavily on implicit background knowledge may be scored poorly. A chain may also not reflect the model's actual internal reasoning. Code: github.com/archiki/ReCEval.

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 North Carolina at Chapel HillChapel Hill · United States37 works · 14,332 citations

📚 Shelves 5The 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 worksComputer Vision and Pattern Recognition · 1,282 works

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

Cite

Cite as Archiki Prasad, Swarnadeep Saha, Xiang Zhou and Mohit Bansal. 2023. ReCEval: Evaluating Reasoning Chains via Correctness and Informativeness. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2023.emnlp-main.622

@inproceedings{prasad2023receval,
  title = {{ReCEval: Evaluating Reasoning Chains via Correctness and Informativeness}},
  author = {Archiki Prasad, Swarnadeep Saha, Xiang Zhou and Mohit Bansal},
  booktitle = {Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing},
  year = {2023},
  doi = {10.18653/v1/2023.emnlp-main.622},
  url = {https://doi.org/10.18653/v1/2023.emnlp-main.622}
}

References

64 references, 37 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Shourya Aggarwal, Divyanshu Mandowara, Vishwajeet Agrawal, Dinesh Khandelwal, Parag Singla, and Dinesh Garg. 2021. Explanations for CommonsenseQA: New Dataset and Models. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Confe…
  2. Satanjeev Banerjee and Alon Lavie. 2005. METEOR: An automatic metric for MT evaluation with improved correlation with human judgments. In Proceedings of the ACL Workshop on Intrinsic and Extrinsic Evaluation Measures for Machine Translation and/or Summarization, pages 65–72, Ann Arbor, Michigan. Ass…
  3. Samuel R. Bowman, Gabor Angeli, Christopher Potts, and Christopher D. Manning. 2015. A large annotated corpus for learning natural language inference. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 632–642, Lisbon, Portugal. Association for Computati… find
  4. Information-theoretic evaluation of free-text rationales. arXiv preprint arXiv:2210.04982. — [arXiv:2210.04982](https://arxiv.org/abs/2210.04982) find
  5. Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Sebastian Gehrmann, et al. 2022. PaLM: Scaling language modeling with pathways. arXiv preprint arXiv:2204.02311.
  6. Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, et al. 2022. Scaling instruction-finetuned language models. arXiv preprint arXiv:2210.11416.
  7. Miruna-Adriana Clinciu, Arash Eshghi, and Helen Hastie. 2021. A study of automatic metrics for the evaluation of natural language explanations. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 2376–2387, Online. Associ… find
  8. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, et al. 2021. Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168.
  9. Antonia Creswell and Murray Shanahan. 2022. Faithful reasoning using large language models. arXiv preprint arXiv:2208.14271. find
  10. Bhavana Dalvi, Peter Jansen, Oyvind Tafjord, Zhengnan Xie, Hannah Smith, Leighanna Pipatanangkura, and Peter Clark. 2021. Explaining answers with entailment trees. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7358–7370, Online and Punta Cana, Domin…
  11. Mingkai Deng, Bowen Tan, Zhengzhong Liu, Eric Xing, and Zhiting Hu. 2021. Compression, transduction, and creation: A unified framework for evaluating natural language generation. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 7580–7605, Online and Pu…
  12. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  13. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,…
  14. Hanjie Chen, Faeze Brahman, Xiang Ren, Yangfeng Ji, Yejin Choi, and Swabha Swayamdipta. 2022. find
  15. Rev: Dheeru Dua, Yizhong Wang, Pradeep Dasigi, Gabriel Stanovsky, Sameer Singh, and Matt Gardner. 2019. find
  16. DROP: A reading comprehension benchmark requiring discrete reasoning over paragraphs. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 2368–2378, Minneapolis, Min…
  17. Nan Duan, Duyu Tang, and Ming Zhou. 2020. Machine reasoning: Technology, dilemma and future. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: Tutorial Abstracts, pages 1–6, Online. Association for Computational Linguistics. find
  18. Kawin Ethayarajh, Yejin Choi, and Swabha Swayamdipta. 2022. Understanding Dataset Difficulty with V-Usable Information. In International Conference on Machine Learning, pages 5988–6008. PMLR. find
  19. Jinlan Fu, See-Kiong Ng, Zhengbao Jiang, and Pengfei Liu. 2023a. GPTScore: Evaluate as you desire. arXiv preprint arXiv:2302.04166. find
  20. Yao Fu, Hao Peng, Ashish Sabharwal, Peter Clark, and Tushar Khot. 2023b. Complexity-based prompting for multi-step reasoning. In The Eleventh International Conference on Learning Representations.
  21. Matt Gardner, Joel Grus, Mark Neumann, Oyvind Tafjord, Pradeep Dasigi, Nelson F. Liu, Matthew Peters, Michael Schmitz, and Luke Zettlemoyer. 2018. AllenNLP: A deep semantic natural language processing platform. In Proceedings of Workshop for NLP Open Source Software (NLP-OSS), pages 1–6, Melbourne,…
  22. Mor Geva, Daniel Khashabi, Elad Segal, Tushar Khot, Dan Roth, and Jonathan Berant. 2021. Did aristotle use a laptop? a question answering benchmark with implicit reasoning strategies. Transactions of the Association for Computational Linguistics, 9:346– 361. find
  23. Olga Golovneva, Moya Chen, Spencer Poff, Martin Corredor, Luke Zettlemoyer, Maryam Fazel-Zarandi, and Asli Celikyilmaz. 2023. ROSCOE: A suite of metrics for scoring step-by-step reasoning. In The Eleventh International Conference on Learning Representations.
  24. Simeng Han, Hailey Schoelkopf, Yilun Zhao, Zhenting Qi, Martin Riddell, Luke Benson, Lucy Sun, Ekaterina Zubova, Yujie Qiao, Matthew Burtell, et al. 2022. FOLIO: Natural language reasoning with first-order logic. arXiv preprint arXiv:2209.00840.
  25. Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, Steven Basart, Eric Tang, Dawn Song, and Jacob Steinhardt. 2021. Measuring mathematical problem solving with the math dataset. Advances in Neural Information Processing Systems.
  26. John Hewitt, Kawin Ethayarajh, Percy Liang, and Christopher Manning. 2021. Conditional probing: measuring usable information beyond a baseline. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 1626–1639, Online and Punta Cana, Dominican Republic. Assoc… find
  27. Jie Huang and Kevin Chen-Chuan Chang. 2022. Towards reasoning in large language models: A survey. arXiv preprint arXiv:2212.10403.
  28. 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…
  29. Albert Qiaochu Jiang, Sean Welleck, Jin Peng Zhou, Timothee Lacroix, Jiacheng Liu, Wenda Li, Mateja Jamnik, Guillaume Lample, and Yuhuai Wu. 2023. Draft, sketch, and prove: Guiding formal theorem provers with informal proofs. In The Eleventh International Conference on Learning Representations.
  30. Saurav Kadavath, Tom Conerly, Amanda Askell, Tom Henighan, Dawn Drain, Ethan Perez, Nicholas Schiefer, Zac Hatfield-Dodds, Nova DasSarma, Eli Tran-Johnson, et al. 2022. Language models (mostly) know what they know. arXiv preprint arXiv:2207.05221. find
  31. Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. Large language models are zero-shot reasoners. In Advances in Neural Information Processing Systems.
  32. Moritz Laurer, W v Atteveldt, Andreu Casas, and Kasper Welbers. 2022. Less annotating, more classifying–addressing the data scarcity issue of supervised machine learning with deep transfer learning and bert-nli. find
  33. Chin-Yew Lin. 2004. ROUGE: A package for automatic evaluation of summaries. In Text Summarization Branches Out, pages 74–81, Barcelona, Spain. Association for Computational Linguistics.
  34. Kevin Lin, Oyvind Tafjord, Peter Clark, and Matt Gardner. 2019. Reasoning over paragraph effects in situations. In Proceedings of the 2nd Workshop on Machine Reading for Question Answering, pages 58–62, Hong Kong, China. Association for Computational Linguistics.
  35. Zhan Ling, Yunhao Fang, Xuanlin Li, Zhiao Huang, Mingu Lee, Roland Memisevic, and Hao Su. 2023. Deductive verification of chain-of-thought reasoning. arXiv preprint arXiv:2306.03872.
  36. Yang Liu, Dan Iter, Yichong Xu, Shuohang Wang, Ruochen Xu, and Chenguang Zhu. 2023. Gpteval: Nlg evaluation using gpt-4 with better human alignment. arXiv preprint arXiv:2303.16634. find
  37. Abulhair Saparov and He He. 2023. Language models can (kind of) reason: A systematic formal analysis of chain-of-thought. In International Conference on Learning Representations. find
  38. Qing Lyu, Shreya Havaldar, Adam Stein, Li Zhang, Delip Rao, Eric Wong, Marianna Apidianaki, and Chris Callison-Burch. 2023. Faithful chain-of-thought reasoning. arXiv preprint arXiv:2301.13379.
  39. Thibault Sellam, Dipanjan Das, and Ankur Parikh. 2020. BLEURT: Learning robust metrics for text generation. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7881–7892, Online. Association for Computational Linguistics. find
  40. Ani Nenkova and Rebecca Passonneau. 2004. Evaluating content selection in summarization: The pyramid method. In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics: HLT-NAACL 2004, pages 145–152, Boston, Massachusetts… find
  41. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730–27744.
  42. Vishakh Padmakumar and He He. 2021. Unsupervised extractive summarization using pointwise mutual information. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 2505–2512, Online. Association for Computational Linguistic… find
  43. Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311–318, Philadelphia, Pennsylvania, USA. Association for Computational…
  44. Fabio Petroni, Tim Rocktäschel, Sebastian Riedel, Patrick Lewis, Anton Bakhtin, Yuxiang Wu, and Alexander Miller. 2019. Language models as knowledge bases? In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural…
  45. Claude E Shannon. 1948. A mathematical theory of communication. The Bell system technical journal, 27(3):379–423. find
  46. Ori Shapira, David Gabay, Yang Gao, Hadar Ronen, Ramakanth Pasunuru, Mohit Bansal, Yael Amsterdamer, and Ido Dagan. 2019. Crowdsourcing lightweight pyramids for manual summary evaluation. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguis… find
  47. Peng Shi and Jimmy Lin. 2019. Simple bert models for relation extraction and semantic role labeling. arXiv preprint arXiv:1904.05255.
  48. Robert H Somers. 1962. A new asymmetric measure of association for ordinal variables. American sociological review, pages 799–811. find
  49. Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2019. CommonsenseQA: A question answering challenge targeting commonsense knowledge. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,…
  50. Brian Thompson and Matt Post. 2020. Automatic machine translation evaluation in many languages via zero-shot paraphrasing. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 90–121, Online. Association for Computational Linguistics.
  51. Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J Liu. 2020. Exploring the limits of transfer learning with a unified text-to-text transformer. The Journal of Machine Learning Research, 21(1):5485–5551.
  52. Jidong Tian, Yitian Li, Wenqing Chen, Liqiang Xiao, Hao He, and Yaohui Jin. 2021. Diagnosing the first-order logical reasoning ability through LogicNLI. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 3738–3747, Online and Punta Cana, Dominican Republ…
  53. Adam Roberts, Colin Raffel, and Noam Shazeer. 2020. How much knowledge can you pack into the parameters of a language model? In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 5418–5426, Online. Association for Computational Linguistics. find
  54. Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971.
  55. Miles Turpin, Julian Michael, Ethan Perez, and Samuel R Bowman. 2023. Language models don’t always say what they think: Unfaithful explanations in chain-of-thought prompting. arXiv preprint arXiv:2305.04388.
  56. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V Le, Ed H. Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2023. Self-consistency improves chain of thought reasoning in language models. In The Eleventh International Conference on Learning Representations.
  57. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, brian ichter, Fei Xia, Ed H. Chi, Quoc V Le, and Denny Zhou. 2022. Chain of thought prompting elicits reasoning in large language models. In Advances in Neural Information Processing Systems.
  58. Sean Welleck, Jiacheng Liu, Ronan Le Bras, Hannaneh Hajishirzi, Yejin Choi, and Kyunghyun Cho. 2021. Naturalproofs: Mathematical theorem proving in natural language. In Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1). find
  59. Sean Welleck, Jiacheng Liu, Ximing Lu, Hannaneh Hajishirzi, and Yejin Choi. 2022. Naturalprover: Grounded mathematical proof generation with language models. In Advances in Neural Information Processing Systems. find
  60. Adina Williams, Nikita Nangia, and Samuel Bowman. 2018. A broad-coverage challenge corpus for sentence understanding through inference. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Pa… find
  61. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, Mariama Drame, Quentin Lho…
  62. Yilun Xu, Shengjia Zhao, Jiaming Song, Russell Stewart, and Stefano Ermon. 2020. A theory of usable information under computational constraints. International Conference on Learning Representations. find
  63. Weizhe Yuan, Graham Neubig, and Pengfei Liu. 2021. BARTScore: Evaluating generated text as text generation. Advances in Neural Information Processing Systems, 34:27263–27277.
  64. Algorithm 1 Chain-level Scores in R E CE VAL find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/prasad-2023-receval-evaluating-reasoning-chains-correctness/