Can Large Language Models Capture Dissenting Human Voices?
This addresses concerns about LLMs' natural language understanding and representativeness for broader human populations, though it is incremental as it builds on existing evaluation methods.
The paper tackled the problem of whether large language models (LLMs) align with human disagreement distributions in natural language inference tasks, finding that LLMs exhibit limited performance and fail to capture human disagreement, with further declines in high-disagreement scenarios.
Large language models (LLMs) have shown impressive achievements in solving a broad range of tasks. Augmented by instruction fine-tuning, LLMs have also been shown to generalize in zero-shot settings as well. However, whether LLMs closely align with the human disagreement distribution has not been well-studied, especially within the scope of natural language inference (NLI). In this paper, we evaluate the performance and alignment of LLM distribution with humans using two different techniques to estimate the multinomial distribution: Monte Carlo Estimation (MCE) and Log Probability Estimation (LPE). As a result, we show LLMs exhibit limited ability in solving NLI tasks and simultaneously fail to capture human disagreement distribution. The inference and human alignment performances plunge even further on data samples with high human disagreement levels, raising concerns about their natural language understanding (NLU) ability and their representativeness to a larger human population. The source code for the experiments is available at https://github.com/xfactlab/emnlp2023-LLM-Disagreement