CollabFlow: Recursive Self-Improvement of Agent Collaboration (Q7360765)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
No description defined
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | CollabFlow: Recursive Self-Improvement of Agent Collaboration |
No description defined |
Statements
CollabFlow: Recursive Self-Improvement of Agent Collaboration (English)
0 references
29 September 2026
0 references
cs.MA
0 references
cs.AI
0 references
Xiao Huang
0 references
Mingda Zhang
0 references
Junming Zhang
0 references
Qiang Huang
0 references
Hanwen Zhang
0 references
Yue Dai
0 references
Zijia Wang
0 references
Xiaoying Tang
0 references
We propose a recursive self-improvement system of learned agent collaboration where a director constructs teams of agents based on their past performance, evidence-conditioned communication protocols are learned in each round, and a flow-based objective encourages the emergence and retention of multiple effective teams.
0 references
Multi-Agent
0 references
Recursive Self Improvement
0 references
Collaborative Learning
0 references
Evidence-Conditioned Communication
0 references
1 reference
