English

AbductionRules: Training Transformers to Explain Unexpected Inputs

Computation and Language 2022-03-24 v1

Abstract

Transformers have recently been shown to be capable of reliably performing logical reasoning over facts and rules expressed in natural language, but abductive reasoning - inference to the best explanation of an unexpected observation - has been underexplored despite significant applications to scientific discovery, common-sense reasoning, and model interpretability. We present AbductionRules, a group of natural language datasets designed to train and test generalisable abduction over natural-language knowledge bases. We use these datasets to finetune pretrained Transformers and discuss their performance, finding that our models learned generalisable abductive techniques but also learned to exploit the structure of our data. Finally, we discuss the viability of this approach to abductive reasoning and ways in which it may be improved in future work.

Keywords

Cite

@article{arxiv.2203.12186,
  title  = {AbductionRules: Training Transformers to Explain Unexpected Inputs},
  author = {Nathan Young and Qiming Bao and Joshua Bensemann and Michael Witbrock},
  journal= {arXiv preprint arXiv:2203.12186},
  year   = {2022}
}

Comments

Findings of ACL 2022

R2 v1 2026-06-24T10:22:54.642Z