CLAIDec 17, 2024

DateLogicQA: Benchmarking Temporal Biases in Large Language Models

arXiv:2412.13377v211 citationsh-index: 7NAACL
Originality Synthesis-oriented
AI Analysis

This work addresses the problem of temporal biases in LLMs for researchers and developers, but it is incremental as it focuses on benchmarking and analysis without proposing a new method.

The paper tackles the problem of temporal reasoning in large language models by introducing DateLogicQA, a benchmark with 190 questions, and finds that models exhibit biases at representation and logical levels, highlighting challenges in handling temporal data accurately.

This paper introduces DateLogicQA, a benchmark with 190 questions covering diverse date formats, temporal contexts, and reasoning types. We propose the Semantic Integrity Metric to assess tokenization quality and analyse two biases: Representation-Level Bias, affecting embeddings, and Logical-Level Bias, influencing reasoning outputs. Our findings provide a comprehensive evaluation of LLMs' capabilities and limitations in temporal reasoning, highlighting key challenges in handling temporal data accurately.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes