NEWS

LLMs

Research on SMARTCAL Accepted at EMNLP 2024 Industry Track

2024.10.15

Congratulations to Yuanhao Shen and Professor Xiaodan Zhu from Queen’s University, and Lei Chen from Rakuten Institute of Technology in Boston on the acceptance of their paper at the EMNLP 2024 Industry Track!

Their paper, titled “SMARTCAL: An Approach to Self-Aware Tool-Use Evaluation and Calibration,” explores the tool-use abilities of Large Language Models (LLMs) and uncovers significant issues with tool misuse and overconfidence. To tackle these challenges, they developed SMARTCAL, a novel framework that has shown an 8.6% increase in QA performance and a 21.6% reduction in Expected Calibration Error (ECE) across three datasets.

EMNLP 2024 will be held from November 12 to 16, 2024, in Miami, Florida!

Copied! instagram