cooperbench reports
Experiment & ablation reports across the org, grouped by source repo. Each file is a self-contained page; published on every push by the reports repo's CI.
cooperbench
—
Team-Harness Ablation & Multi-Agent Comparison
LATEST
2026-06-08
CooperAgents — Unified Self-Evolving Harness
2026-05-29
qwen35-9b · team-no-protocol · CooperData v1
2026-05-26
Team → Coop Dataset — placeholder
2026-05-26
Coordination Harness Pareto — temporary report
2026-05-23
Team Trajectory Viewer
2026-05-23
CooperBench Coordination Study — full dataset
coopertrain
2026-05-19
Team-Harness Ablation & Multi-Agent Comparison
LATEST
2026-05-10
SFT + TITO distillation for plan-first coop agents — teacher-bounded follow-through
2026-05-08
Team experiment registry — snapshot 2026-05-08
swe-chat
2026-06-12
SWE-chat: Users & Trajectories
LATEST
user-skill
2026-06-16
User.skill: how faithfully can we simulate a developer?
LATEST
2026-06-16
Complete session: real vs simulated developer (pavel401)
2026-06-16
Complete session: real vs simulated developer (marcus-sa)
2026-06-14
User.skill: how faithfully can we simulate a developer?
2026-06-13
User.skill: recognizable vs. realistic role-play