Can we trust the evaluation on ChatGPT?
Authors:
Preprint
Abstract
ChatGPT, the first large language model (LLM) with mass adoption, has demonstrated remarkable performance in numerous natural language tasks. Despite its evident usefulness, evaluating ChatGPT's performance in diverse problem domains remains challenging due to the closed nature of the model and its continuous updates via Reinforcement Learning from Human Feedback (RLHF). We highlight the issue of data contamination in ChatGPT evaluations, with a case study of the task of stance detection. We discuss the challenge of preventing data contamination and ensuring fair model evaluation in the age of closed and continuously trained models.Keywords
All Related Versions
This publication has 7 references indexed in Scilit:
- ChatGPT: Jack of all trades, master of noneInformation Fusion, 2023
- An Analysis of the Automatic Bug Fixing Performance of ChatGPT2023
- “What Can ChatGPT Do?” Analyzing Early Reactions to the Innovative AI Chatbot on TwitterBig Data and Cognitive Computing, 2023
- ChatGPT for (Finance) research: The Bananarama ConjectureFinance Research Letters, 2023
- Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?2022
- Documenting Large Webtext Corpora: A Case Study on the Colossal Clean Crawled Corpus2021
- Catastrophic Interference in Connectionist Networks: The Sequential Learning Problem1989