data-pollution

#data-pollution

@huangyihe: When joining Tencent as a consultant, Yao Shunyu had only one task: to investigate the reasons behind the long-term lag of the Hunyuan large model. The finding was: "In simple terms, almost every link is leaking." Hunyuan excessively pursued leaderboard results, contaminating the training data with test-prep materials. The model became good at exams but performed poorly in real-world scenarios. At that time, the accuracy acceptance threshold for data annotation was set at 95%, but in reality it remained at 60%-70% for a long time. Teams produced large amounts of unusable data just to meet deadlines, and the algorithm team tacitly allowed this.

X AI KOLs Timeline ↗ · 2026-07-14 Cached

Tencent consultant Yao Shunyu investigated the lagging performance of the Hunyuan large model, finding systemic issues such as data contamination from chasing leaderboard scores and poor annotation quality.

0 favorites 0 likes

data-pollution

Submit Feedback