Stanford CS329X HW3: Collecting Human-AI Interaction Data with Coco
Stanford's CS329X course released HW3, where students use a simplified fork of the open-source agent Coco to simulate and collect multi-turn human-agent interaction trajectories, then build evaluation metrics with AutoMetrics to assess human-AI collaboration.
2026-11-04 ~ 2026-11-04 · 2 related posts
- Stanford CS329X HW3: Coco agent traces plus AutoMetrics evaluators, open-sourced — EchoShao8899 · 2026-11-04
- Stanford CS329X HW3 has students simulate human-agent interaction and build auto-evaluators — EchoShao8899 · 2026-11-04