mlsanityData to 7 Oct 2026

Generating Intermediate Steps for NLI with Next-Step Supervision

arXiv 2208.14641 ↗First posted 31 Aug 2022cs.CL cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

The Natural Language Inference (NLI) task often requires reasoning over multiple steps to reach the conclusion. While the necessity of generating such intermediate steps (instead of a summary explanation) has gained popular support, it is unclear how to generate such steps without complete end-to-end supervision and how such generated steps can be further utilized. In this work, we train a sequence-to-sequence model to generate only the next step given an NLI premise and hypothesis pair (and previous steps); then enhance it with external knowledge and symbolic search to generate intermediate steps with only next-step supervision. We show the correctness of such generated steps through automated and human verification. Furthermore, we show that such generated steps can help improve end-to-end NLI task performance using simple data augmentation strategies, across multiple public NLI datasets.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 27 of its 34 references that are in the map. First posted on arXiv 31 Aug 2022.

Builds on

27 of 34 references.

References · 27

  1. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  3. 2017A Broad-Coverage Challenge Corpus for Sentence Understanding through Inferencecited by 2,665
  4. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,126
  5. 2015A large annotated corpus for learning natural language inferencecited by 1,837
  6. 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,219
  7. 2020Beyond Accuracy: Behavioral Testing of NLP models with CheckListcited by 979
  8. 2018Adafactor: Adaptive Learning Rates with Sublinear Memory Costcited by 896
  9. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 871
  10. 2016The Mythos of Model Interpretabilitycited by 716
  11. 2018e-SNLI: Natural Language Inference with Natural Language Explanationscited by 468
  12. 2019Learning the Difference that Makes a Difference with Counterfactually-Augmented Datacited by 428
15 more
  1. 2020ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Languagecited by 309
  2. 2020Transformers as Soft Reasoners over Languagecited by 265
  3. 2021Explaining Answers with Entailment Treescited by 148
  4. 2021Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processingcited by 126
  5. 2021Improving Multi-hop Knowledge Base Question Answering by Learning Intermediate Supervision Signalscited by 100
  6. 2020NILE : Natural Language Inference with Faithful Natural Language Explanationscited by 96
  7. 2019Graph Representations for Higher-Order Logic and Theorem Provingcited by 55
  8. 2020PRover: Proof Generation for Interpretable Reasoning over Rulescited by 53
  9. 2020GenericsKB: A Knowledge Base of Generic Statementscited by 51
  10. 2019Can neural networks understand monotonicity reasoning?cited by 45
  11. 2020Measuring Systematic Generalization in Neural Proof Generation with Transformerscited by 43
  12. 2020TaxiNLI: Taking a Ride up the NLU Hillcited by 27
  13. 2020Learning to Prove from Synthetic Theoremscited by 13
  14. 2021Analyzing the Nuances of Transformers' Polynomial Simplification Abilitiescited by 9
  15. 2021Trusting RoBERTa over BERT: Insights from CheckListing the Natural Language Inference Taskcited by 8

Built on by

Not cited yet.