TeamLLM improves LLM performance on multi-step contextualized tasks by using four human-like team roles and a three-phase collaboration process, demonstrated on the new CGPST benchmark.
• Phase B – Pilot Demonstration (Nov 2025–Apr 2026): Deploy 100 kg in South Bay, Oahu; 30-day monitoring shows≥70%reduction of microplastics in plankton
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
TeamLLM: A Human-Like Team-Oriented Collaboration Framework for Multi-Step Contextualized Tasks
TeamLLM improves LLM performance on multi-step contextualized tasks by using four human-like team roles and a three-phase collaboration process, demonstrated on the new CGPST benchmark.