Benchmarking and Confidence Evaluation of LALMs For Temporal Reasoning
This work addresses the need for better benchmarking of LALMs in reasoning tasks, which is crucial for the multimodal AI community, though it is incremental as it builds on existing evaluation frameworks.
The authors tackled the problem of evaluating large audio language models (LALMs) on temporal reasoning tasks by introducing a new dataset called TREA, and found that LALMs consistently underperform compared to humans on these tasks. They also proposed an uncertainty metric that shows no correlation with accuracy, highlighting the need for more comprehensive evaluation in high-stakes applications.
The popular success of text-based large language models (LLM) has streamlined the attention of the multimodal community to combine other modalities like vision and audio along with text to achieve similar multimodal capabilities. In this quest, large audio language models (LALMs) have to be evaluated on reasoning related tasks which are different from traditional classification or generation tasks. Towards this goal, we propose a novel dataset called temporal reasoning evaluation of audio (TREA). We benchmark open-source LALMs and observe that they are consistently behind human capabilities on the tasks in the TREA dataset. While evaluating LALMs, we also propose an uncertainty metric, which computes the invariance of the model to semantically identical perturbations of the input. Our analysis shows that the accuracy and uncertainty metrics are not necessarily correlated and thus, points to a need for wholesome evaluation of LALMs for high-stakes applications.