Foundational Automatic Evaluators: Scaling Multi-Task Generative Evaluator Training for Reasoning-Centric Domains
Paper
•
2510.17793
•
Published
•
2
FARE
FARE are Salesforce AI Research's open multi-task evaluator models.