A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity

By Yejin Bang, Samuel Cahyawijaya, Nayeon Lee, Wenliang Dai, Dan Su, Bryan Wilie, Holy Lovenia, Ziwei Ji, Tiezheng Yu, Willy Chung and 3 more. International Joint Conference on Natural Language Processing, 2023

1,776 citations recorded by the source corpus.

This paper proposes a framework for quantitatively evaluating interactive LLMs such as ChatGPT using publicly available data sets. We carry out an extensive technical evaluation of ChatGPT using 23 data sets covering 8 different common NLP application tasks. We evaluate the multitask, multilingual and multi-modal aspects of ChatGPT based on these data sets and a newly designed multimodal dataset. We find that ChatGPT outperforms LLMs with zero-shot learning on most tasks and even outperforms fine-tuned models on some tasks. We find that it is better at understanding non-Latin script languages than generating them. It is able to generate multimodal content from textual prompts, via an intermediate code generation step. Moreover, we find that ChatGPT is 63.41% accurate on average in 10 different reasoning categories under logical reasoning, non-textual reasoning, and commonsense reasoning, hence making it an unreliable reasoner. It is, for example, better at deductive than inductive reasoning. ChatGPT suffers from hallucination problems like other LLMs and it generates more extrinsic hallucinations from its parametric memory as it does not have access to an external knowledge base. Finally, the interactive feature of ChatGPT enables human collaboration with the underlying LLM to improve its performance, i.e, 8% ROUGE-1 on summarization and 2% ChrF++ on machine translation, in a multi-turn"prompt engineering"fashion. We also release codebase for evaluation set extraction.

Subfields: Applied Psychology, Health Informatics, Social Psychology

View Paper: https://www.semanticscholar.org/paper/bf8491bef353df126e2306ad2fe4b898697b906a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Applied Psychology

  1. Ashwin Kalyan
  2. Dragan Gasevic

In Health Informatics

  1. Heng Ji
  2. Matei A. Zaharia
  3. Alexander J. Ratner
  4. Siyuan Ma
  5. James Y. Zou
  6. Zheng-Xin Yong
  7. Mihai Surdeanu
  8. Nikolaos D. Tselikas

In Social Psychology

  1. James Y. Zou
  2. Ranjay Krishna
  3. Mihai Surdeanu
  4. Diyi Yang
  5. Dragan Gasevic

Citations in the corpus