The authors separate an agent-improvement objective from the procedure used to pursue it.
Agent evolvers can learn their own search strategies across tasks
FreeEvolve lets an agent control and refine its optimization campaign, replacing fixed human-designed search loops with a reusable learned policy.
Big Tech
Lecheng Kong · Like Hui · Nikos Kanakaris · Prithwish Jana · Sahika Genc · Narayanan Sadagopan
AWS AI Labs · Georgia Institute of Technology
Research Digest··2 min read
Kong et al.
Why this paper
From AWS AI Labs and Georgia Institute of Technology
In one line
A learned campaign policy lets agent evolvers organize their own search and match or exceed hand-designed loops while transferring across tasks.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ✓Limitations stated by the authors (2 noted)
- ✓Reports numbers on named benchmarks (4 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§