Human-Computer QA Competition: June 14 & 21

Participate in the Human-Computer Question Answering Competition on June 14 & 21. Evaluate language models on accuracy and uncertainty. Beginner-friendly submission process. More info at https://sites.google.com/view/qanta/2025-competition

A new competition is announced for machine learning and NLP enthusiasts. The Human-Computer Question Answering Competition will take place on June 14 (in-person, DC) and June 21 (online). The competition aims to evaluate language models not just on accuracy, but also on their ability to communicate uncertainty.

The primary metric will be how well models help users answer questions. A model that is 66% accurate but correctly identifies its limitations will be considered better than a 75% accurate model that is convincingly wrong on the remaining 25%.

• Models will be evaluated on accuracy and ability to convey uncertainty.
• The competition is designed to be beginner-friendly, with options for prompt-based and huggingface model uploads.
• The human-computer games will be filmed and posted on YouTube.

For more information, visit https://sites.google.com/view/qanta/2025-competition. Contact qanta@googlegroups.com for questions.

Tags: NLP, Machine Learning, Human-Computer Interaction, Question Answering, AI Competition, Language Models, Uncertainty in AI