Paper, audited contrast datasets, and interactive workbench for Parity-Aware Compute Evaluation (arXiv:2611.17610).
-
More Agents, More Samples, or More Thinking? Untangling Coordination, Ensembling, and Test-Time Compute in Multi-Agent LLMs
Paper • 2611.17610 • Published • 4 -
sajjadi/pace-contrasts
Viewer • Updated • 1.36k • 36 -
sajjadi/hidden-profile-control
Viewer • Updated • 500 • 25 -
PACE Studio
🚀1Analyze multi-agent LLM traces and compute baseline metrics