Researchers release ActiveSaddler
ActiveSaddler released — points out that existing LLM-agent optimization deals with "how to update the harness" but largely leaves "which training scenarios produce that feedback" fixed, and adapts the curriculum together with harness evolution using a non-stationary bandit. It abstracts recurring agent failures into reusable failure patterns and estimates the learning potential of targeting each pattern, weighing exploitation of known weaknesses against exploration of new failure situations. Compared with baseline harness optimization using a fixed scenario order, +4.4 percentage points Pass@1 on GAIA2 and +7.5 points on Terminal-Bench 2.0. arXiv 2610.00906.
Why it matters
The optimization target of harness engineering has widened from "what to fix" to "what to test". Abstracting failures into patterns to compose the curriculum automatically takes over the place where people used to pick scenarios in the agent-improvement loop, and touches on recursive self-improvement (axis ⑫).
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/abs/2610.00906Added in Week 41, 2026 · See that week