Kangwook Lee: agent harness improvement can happen inside rollouts once agents are capable enough

Kangwook_Lee · x · 2026-09-04

Responding to a question about his paper, Kangwook Lee explains that the distinction between a "meta-harness" (the one that trains) and "harness" (the one being trained) was a deliberate setup choice, not a necessity.

Once that distinction goes away and agents become capable enough, harness improvement can happen as part of the rollouts themselves — in-context harness adaptation.

Related event: KRAFTON Open-Sources WHALE: Alternating Weight and Harness Optimization Boosts LLM Agents(8 posts)→

Original post →

More from Research

Research channel →