# Learning to Rank for Plausible Plausibility

arXiv: https://arxiv.org/abs/1906.02079 · HTML: https://mlsanity.com/paper/1906.02079/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Jun 2019.

## Built on by (most cited first, 9 of 9)

- [Guided Generation of Cause and Effect](https://mlsanity.com/paper/2107.09846/) (2021)
- [Are Machines Better at Complex Reasoning? Unveiling Human-Machine Inference Gaps in Entailment Verification](https://mlsanity.com/paper/2402.03686/) (2024)
- [CausalBERT: Injecting Causal Knowledge Into Pre-trained Models with Minimal Supervision](https://mlsanity.com/paper/2107.09852/) (2021)
- [Improving Commonsense Causal Reasoning by Adversarial Training and Data Augmentation](https://mlsanity.com/paper/2101.04966/) (2021)
- [Always Tell Me The Odds: Fine-grained Conditional Probability Estimation](https://mlsanity.com/paper/2505.01595/) (2025)
- [Knowledge-Augmented Language Models for Cause-Effect Relation Classification](https://mlsanity.com/paper/2112.08615/) (2021)
- [ReCo: Reliable Causal Chain Reasoning via Structural Causal Recurrent Neural Networks](https://mlsanity.com/paper/2212.08322/) (2022)
- [Advancing Automated Spatio-Semantic Analysis in Picture Description Using Language Models](https://mlsanity.com/paper/2510.05128/) (2025)
- [Modelling Adjectival Modification Effects on Semantic Plausibility](https://mlsanity.com/paper/2507.21828/) (2025)

## Abstract (arXiv)

Researchers illustrate improvements in contextual encoding strategies via resultant performance on a battery of shared Natural Language Understanding (NLU) tasks. Many of these tasks are of a categorical prediction variety: given a conditioning context (e.g., an NLI premise), provide a label based on an associated prompt (e.g., an NLI hypothesis). The categorical nature of these tasks has led to common use of a cross entropy log-loss objective during training. We suggest this loss is intuitively wrong when applied to plausibility tasks, where the prompt by design is neither categorically entailed nor contradictory given the context. Log-loss naturally drives models to assign scores near 0.0 or 1.0, in contrast to our proposed use of a margin-based loss. Following a discussion of our intuition, we describe a confirmation study based on an extreme, synthetically curated task derived from MultiNLI. We find that a margin-based loss leads to a more plausible model of plausibility. Finally, we illustrate improvements on the Choice Of Plausible Alternative (COPA) task through this change in loss.