FDE Gym contribution
Can a delivered process keep affecting later cases after its own evaluation ends?
A playground of connected ideas.
Explore the questions connecting public research, open-source tools, and applications.
Replay strategies
When does a replay strategy hold up across different random seeds?
TSAD evaluation
How much of a disagreement between anomaly-detection metrics is meaningful?
jevcompat
How can an API compatibility claim become something another person can check?
SessionReel
How can a coding-agent session become an artifact someone else can inspect?
Second Look
When a model changes, which old work is worth revisiting?
Equilaw
How can a multi-step AI workflow make its intermediate results understandable?