RAO uses RL to train recursive agents that delegate sub-tasks to self-copies, yielding better training efficiency, generalization to harder tasks, scaling beyond context windows, and lower wall-clock time.
m0_i2": 4}, 20) - Parallel: results = await asyncio.gather( launch_subagent({
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Recursive Agent Optimization
RAO uses RL to train recursive agents that delegate sub-tasks to self-copies, yielding better training efficiency, generalization to harder tasks, scaling beyond context windows, and lower wall-clock time.