有人說前麵提到月之暗麵的研究結果早有人發過類似論文，貢獻評估還是得過會議雙盲審才可靠；不管怎樣，即使是重新發現也引人注目

https://x.com/behrouz_ali/status/2033581834953453853

https://bbs.wenxuecity.com/znjy/7800831.html

also:

Zhilin Yang (founder Moonshot) is speaking at GTC on scaling Kimi 2.5.

Time: Tuesday, March 17, 11:00–11:40 a.m. PDT (GTC 2026 in San Jose).

He is one of the coauthors of the original T5 paper, “Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer”, cmu PhD '19

https://x.com/bookwormengr/status/2033527837333086559

https://kimiyoung.github.io/