MARA combines flow-matching loss prediction with a multi-agent transformer allocator to improve task completion rates in discrete-node computational resource efficient learning.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MARA: Flow-Matching-Guided Multi-Agent Resource Allocation for Computational Resource Efficient Learning
MARA combines flow-matching loss prediction with a multi-agent transformer allocator to improve task completion rates in discrete-node computational resource efficient learning.