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

By Lianhui Qin, Vered Shwartz, Peter West, Chandra Bhagavatula, Jena D. Hwang, Ronan Le Bras, Antoine Bosselut, Yejin Choi. 2020

96 citations recorded by the source corpus.

Abductive and counterfactual reasoning, core abilities of everyday human cognition, require reasoning about what might have happened at time t, while conditioning on multiple contexts from the relative past and future. However, simultaneous incorporation of past and future contexts using generative language models (LMs) can be challenging, as they are trained either to condition only on the past context or to perform narrowly scoped text-infilling. In this paper, we propose DeLorean, a new unsupervised decoding algorithm that can flexibly incorporate both the past and future contexts using only off-the-shelf, left-to-right language models and no supervision. The key intuition of our algorithm is incorporating the future through back-propagation, during which, we only update the internal representation of the output while fixing the model parameters. By alternating between forward and backward propagation, DeLorean can decode the output representation that reflects both the left and right contexts. We demonstrate that our approach is general and applicable to two nonmonotonic reasoning tasks: abductive text generation and counterfactual story revision, where DeLorean outperforms a range of unsupervised and some supervised methods, based on automatic and human evaluation.

Subfields: Artificial Intelligence, Natural Language Processing

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

Citations in the corpus

Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey (2022, methodology): “These algorithms incorporate control using Monte Carlo Markov Chain (MCMC) techniques (Hoang et al., 2017; Qin et al., 2020; Mireshghallah et al., 2022), in which a full sequence is initialized and iteratively updated.”
Diffusion-LM Improves Controllable Text Generation (2022, methodology): “Table 5: For sentence infilling, Diffusion-LM significantly outperforms prior work COLD [37] and Delorean [36] (numbers taken from paper), and matches the performance of an autoregressive LM (AR) trained from scratch to do infilling.”
Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task (2022, background): “It is related to the ideas behind plug-and-play controllable text generation for autoregressive (Dathathri et al., 2019; Qin et al., 2020; Krause et al., 2020) and diffusion (Li et al., 2022) language models by optimizing the likelihood of the desired attribute and the fluency of generated texts at…”
The Abduction of Sherlock Holmes: A Dataset for Visual Abductive Reasoning (2022, background): “Abductive reasoning has recently received attention in language processing tasks [6, 11, 45, 50], proof writing [60], and dis-”
G ENIE A Leaderboard for Human-in-the-Loop Evaluation of Text Generation (2021, methodology): “For α NLG, we evaluate an unsupervised baseline (Qin et al., 2020) based on GPT-2 (Radford et al., 2019).”
COLD Decoding: Energy-based Constrained Text Generation with Langevin Dynamics (2022, unknown): “1, we compare with DELOREAN [45], a recent state-of-the-art decoding algorithm.”
Gradient-based Constrained Sampling from Language Models (2022, methodology): “After performing multiple steps of this gradient descent, discrete text can be extracted from ˜ y using different heuristics (Kumar et al., 2021; Qin et al., 2020; Song et al., 2020).”
Controlled Text Generation as Continuous Optimization with Multiple Constraints (2021, methodology): “(4)Unlike prior work [19, 50, 59], we do not feed ỹi directly to the model as in our early experiments we found that it leads to slow convergence.”
The Perils of Using Mechanical Turk to Evaluate Open-Ended Text Generation (2021, background): “Each paper was published between 2018 and 2020 at ACL, NAACL, or EMNLP, and we exclude papers that use AMT to evaluate more wellestablished generation tasks like machine translation, or summarization.3 Unlike previous surveys of evaluating generated text (Çelikyilmaz et al., 2020; van der Lee et al.”
Progressive Generation of Long Text with Pretrained Language Models (2021, background): “, 2020), infilling (Zhu et al., 2019; Shen et al., 2020; Qin et al., 2020), or refinement (Lee et al.”