MIT Critical Data

AGORA Phase 0 — a controlled, replayable multi-agent workshop research instrument

How AI agents cheat together, and how to catch them: shortcut cascades and referee oversight in clinical AI. MICCAI 2026.

A unified framework to detect, benchmark, and mitigate shortcut learning in machine learning models across medical imaging and beyond.

An open, community-driven platform for continuous adversarial evaluation of clinical AI, led by MIT Critical Data.

Bridging, Open, Discerning, Humble, Inquiring

Datathon MIT Critical Data Dublin - Ireland

A framework for per-modality failure analysis and missing-data evaluation in multimodal clinical AI.

Literature search pipeline for the creativity-survey review