The authors ran seven agent-evolution methods (prompt, skill, memory and harness evolvers, including GEPA and Meta-Harness) on six models: 1,030 independent evolution runs and over 2.1 million test trajectories. On a pool of 100 candidate…
In the News: November 21, 2026 (Midday)
A reproducibility study finds common ways of choosing an agent's self-written skill edits pick the better one about half the time.
Midday edition
Machine-readable
Download Markdown
Story
Self-improving agents pick the better skill edit about half the time
Read story →
Story
Meta's ML engineering agents kept a change only if it beat a noise probe
The team built agents that propose a change to a search ranker as code, launch a 64-GPU training job, read the offline metrics and decide whether to keep it. Each iteration trained a nominal seven candidates plus a variance probe in…
Story
LangChain describes a "knowledge factory" for agents
Chase separates context, what an agent has during a task, from knowledge, "the reusable understanding we maintain" so the task is done correctly. He presents the idea as the knowledge-work counterpart of the software factory and splits…