Yifan Jiang


2023

pdf bib
BRAINTEASER: Lateral Thinking Puzzles for Large Language Models
Yifan Jiang | Filip Ilievski | Kaixin Ma | Zhivar Sourati
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

The success of language models has inspired the NLP community to attend to tasks that require implicit and complex reasoning, relying on human-like commonsense mechanisms. While such vertical thinking tasks have been relatively popular, lateral thinking puzzles have received little attention. To bridge this gap, we devise BrainTeaser: a multiple-choice Question Answering task designed to test the model’s ability to exhibit lateral thinking and defy default commonsense associations. We design a three-step procedure for creating the first lateral thinking benchmark, consisting of data collection, distractor generation, and generation of adversarial examples, leading to 1,100 puzzles with high-quality annotations. To assess the consistency of lateral reasoning by models, we enrich BrainTeaser based on a semantic and contextual reconstruction of its questions. Our experiments with state-of-the-art instruction- and commonsense language models reveal a significant gap between human and model performance, which is further widened when consistency across adversarial formats is considered. We make all of our code and data available to stimulate work on developing and evaluating lateral thinking models.

2022

pdf bib
Testing Pre-trained Language Models’ Understanding of Distributivity via Causal Mediation Analysis
Pangbo Ban | Yifan Jiang | Tianran Liu | Shane Steinert-Threlkeld
Proceedings of the Fifth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP

To what extent do pre-trained language models grasp semantic knowledge regarding the phenomenon of distributivity? In this paper, we introduce DistNLI, a new diagnostic dataset for natural language inference that targets the semantic difference arising from distributivity, and employ the causal mediation analysis framework to quantify the model behavior and explore the underlying mechanism in this semantically-related task. We find that the extent of models’ understanding is associated with model size and vocabulary size. We also provide insights into how models encode such high-level semantic knowledge.