The authors developed ActiveSaddler for optimizing agent harnesses, the prompts, tool interfaces and control logic surrounding a language model.
Adaptive training curricula produce stronger optimized agent harnesses
ActiveSaddler selects training scenarios according to an agent’s evolving failure patterns, outperforming a fixed scenario order on two benchmarks.
Big Tech
Sungho Park · Wonjoong Kim · Jue Zhang · Wook-Shin Han · Pengfei Gao · Chanyoung Park · +5 more
POSTECH · KAIST · Microsoft
Research Digest··3 min read
Park et al.
Why this paper
From Microsoft and 2 others
In one line
ActiveSaddler adaptively selects training scenarios for harness optimization, raising Pass@1 by 4.4 and 7.5 points on GAIA2 and Terminal-Bench 2.0.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ✓Reports numbers on named benchmarks (2 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§