# Augmenting Lateral Thinking in Language Models with Humor and Riddle Data for the BRAINTEASER Task

arXiv: https://arxiv.org/abs/2405.10385 · HTML: https://mlsanity.com/paper/2405.10385/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning.
- Builds on 5 of its 17 references that are in the map.
- First posted on arXiv 16 May 2024.

## Placed in

- LLM reasoning / [Prompting and chain-of-thought reasoning](https://mlsanity.com/lr/prompting-and-chain-of-thought-reasoning/)

## Builds on (most cited first, 5 of 5)

- [Scaling Language Models: Methods, Analysis & Insights from Training Gopher](https://mlsanity.com/paper/2112.11446/) (2021)
- [SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference](https://mlsanity.com/paper/1808.05326/) (2018)
- [Cosmos QA: Machine Reading Comprehension with Contextual Commonsense Reasoning](https://mlsanity.com/paper/1909.00277/) (2019)
- [OlaGPT: Empowering LLMs With Human-like Problem-Solving Abilities](https://mlsanity.com/paper/2305.16334/) (2023)
- [CODAH: An Adversarially Authored Question-Answer Dataset for Common Sense](https://mlsanity.com/paper/1904.04365/) (2019)

## Abstract (arXiv)

The SemEval 2024 BRAINTEASER task challenges language models to perform lateral thinking -- a form of creative, non-linear reasoning that remains underexplored in NLP. The task comprises two subtasks, Sentence Puzzle and Word Puzzle, requiring models to defy conventional commonsense associations. We present a system that fine-tunes DeBERTaV3 using HuggingFace's AutoModelForMultipleChoice architecture. We augment the provided training data with two additional sources: (1) a humor-style question-answering dataset generated via GPT-4 prompting, and (2) the RiddleSense dataset. This data augmentation strategy is motivated by the observation that humor and riddles share the lateral reasoning structure required by the task. Our best system achieves 92.5% overall accuracy on the Sentence Puzzle subtask and 80.2% on the Word Puzzle subtask, ranking 6th out of 31 teams and 10th out of 23 teams, respectively. We further show that the choice of task formulation matters: framing the problem as multiple-choice rather than sequence classification yields a 10-point accuracy improvement with the same base model. Our analysis reveals that data augmentation with humor and riddle data is particularly effective for sentence-level lateral reasoning, while word-level puzzles remain a harder challenge.