Test-time communication emerges as a new scaling axis: collaborating agent teams beat Best-of-N
DimitrisPapail · x · 2026-09-21
Dimitris Papail and collaborators released a new paper arguing that test-time communication may be the next axis for scaling capabilities, inspired by the Hugging Face incident where agents exploited any channel they could find.
They ran N identical agents on the same task with no prescribed roles, using only a shared text log and instructions to "collaborate". The core question: is Team-of-N better than Best-of-N, when, and why?
Across three research-style tasks, communicating teams decisively outperformed independent agents.
More from coding & agent
- Exa MCP hits 5,000 GitHub stars as AI agents flock to its search integration — TheIshanGoswami · 2026-09-22
- 670,000 agent skills, no trust layer: bot scan finds 69% never reliably fire — markjeffrey · 2026-09-22
- Training on production traces: single-trajectory RL may unlock continual learning — rhythmrg · 2026-09-22
- Anthropic's Swiss cheese model explains why passing evals isn't enough for agents — hugobowne · 2026-09-22
- OpenAI's artists are now all using Codex in their workflow — andrew_n_carr · 2026-09-22
- TinyTorch: PyTorch's free curriculum to build an ML framework from scratch in 20 modules — PyTorch · 2026-09-22