# EvoPilot preprint reports verified gains from agent-led video-search experiments

_Published Monday, September 21, 2026 at 1:09 PM EDT · AI, Science · Latest · Tier 2 — Notable_

Researchers report that EvoPilot, a human-gated system for agent-led model development, found a 3.20-percentage-point offline retrieval improvement during a 37-day campaign on an index of hundreds of millions of videos. A separate seven-day randomized evaluation estimated a 0.66% relative increase in a retention-oriented search-quality measure.

The method assigns specialized agents to experiments while deterministic checks and durable records reject invalid comparisons. Those controls exposed a pre-existing evaluation defect that an earlier automated attempt had wrongly blamed for a 22-percentage-point decline. The results are presented in a preprint and concern one video-discovery system.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2609.21257)

---
Canonical: https://techandbusiness.org/newswire/UtZMNmJGkL2TpE0nSoq2tU
Published: 2026-09-21T17:09:32.790Z
Story chronology: 2026-09-21T04:00:00.000Z
Retrieved: 2026-09-21T19:12:01.008Z
Publisher: Tech & Business (techandbusiness.org)