Learning to Rationalize for Nonmonotonic Reasoning with Distant Supervision

By Faeze Brahman, Vered Shwartz, Rachel Rudinger, Yejin Choi. 2020

46 citations recorded by the source corpus.

The black-box nature of neural models has motivated a line of research that aims to generate natural language rationales to explain why a model made certain predictions. Such rationale generation models, to date, have been trained on dataset-specific crowdsourced rationales, but this approach is costly and is not generalizable to new tasks and domains. In this paper, we investigate the extent to which neural models can reason about natural language rationales that explain model predictions, relying only on distant supervision with no additional annotation cost for human-written rationales. We investigate multiple ways to automatically generate rationales using pre-trained language models, neural knowledge models, and distant supervision from related tasks, and train generative models capable of composing explanatory rationales for unseen instances. We demonstrate our approach on the defeasible inference task, a nonmonotonic reasoning task in which an inference may be strengthened or weakened when new information (an update) is introduced. Our model shows promises at generating post-hoc rationales explaining why an inference is more or less likely given the additional information, however, it mostly generates trivial rationales reflecting the fundamental limitations of neural language models. Conversely, the more realistic setup of jointly predicting the update or its type and generating rationale is more challenging, suggesting an important future direction.

Subfields: Artificial Intelligence, Natural Language Processing, Prompting & Reasoning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

In Prompting & Reasoning

Citations in the corpus

Reframing Human-AI Collaboration for Generating Free-Text Explanations (2021, neutral): “We hope our results can guide future work on free-text explanations via neural or neuro-symbolic systems (Brahman et al., 2021; Majumder et al., 2021; Saha et al., 2021).”
Few-Shot Self-Rationalization with Natural Language Prompts (2021, background): “Brahman et al. (2021) propose a distant supervision approach to explaining a defeasible inference task.”
Thinking Aloud: Dynamic Context Generation Improves Zero-Shot Reasoning Performance of GPT-2 (2021, methodology): “…[Kassner and Schütze, 2020, Richardson and Sabharwal, 2019, Richardson et al., 2020], and can be ﬁne-tuned to correctly draw complex deductive inferences [Clark et al., 2020, Betz et al., 2020] and to generate informal reasons [Rudinger et al., 2020, Camburu et al., 2018, Brahman et al., 2020].”
Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processing (2021, methodology): “[18] that remain after filtering? Are all template-like explanations uninformative?”
Knowledge-Grounded Self-Rationalization via Extractive and Natural Language Explanations (2021, background): “However, NLEs are sometimes produced separately from predictions (Marasovic et al., 2020; Brahman et al., 2021; Atanasova et al., 2020), which raises questions about their faithfulness.”
REV: Information-Theoretic Evaluation of Free-Text Rationales (2022, background): “Chain of Thought Rationales Wei et al. (2022) propose chain of thought prompting to teach large language models to produce intermediate reasoning steps (rationales) before prediction, which improves their prediction performance on a range of reasoning tasks (e.g., arithmetic and symbolic reasoning).”
Teach Me to Explain: A Review of Datasets for Explainable NLP (2021, methodology): “This turned out to be the case with natural language inference (Camburu et al., 2020, also re-ported in Brahman et al., 2021): Explanations in E-SNLI largely follow a set of label-speciﬁc templates.”
Maieutic Prompting: Logically Consistent Reasoning with Recursive Explanations (2022, background): “Belief w E corresponds to the LM’s belief that the proposition E is true (and therefore, ¬ E is false).”
On the Diversity and Limits of Human Explanations (2021, background): “We cannot infer the exact question Brahman et al. (2020) natural lan- guage inference automatically generated.”
Are Hard Examples also Harder to Explain? A Study with Human and Model-Generated Explanations (2022, background): “There has been signiﬁcant progress made in recent years on both curating natural language explanation datasets (Camburu et al., 2018; Rajani et al., 2019; Brahman et al., 2021; Aggarwal et al., 2021, inter alia ) as well as generating them (Rajani et al., 2019; Shwartz et al., 2020).”
ExplaGraphs: An Explanation Graph Generation Task for Structured Commonsense Reasoning (2021, background): “…millions desire food” forms a “V-structure”), (2) unlike natural language explanations (Camburu et al., 2018; Rajani et al., 2019; Narang et al., 2020; Brahman et al., 2021; Zhang et al., 2020), it is easier to impose task-speciﬁc constraints on graphs (e.g., connectivity, acyclicity), that…”