Event ·Multi-source

AI21 Labs 在 NVIDIA GTC Berlin 分享后训练实践与客户案例结果

AI21 Labs CTO Daniel Gissin 在柏林 NVIDIA GTC 的 Post-Training Gathering 上讲解公司的后训练（post-training）思路。内容基于近期与 alice_dot_io、Gong_io、deel 三个客户的合作结果。核心观点是：对开源模型做后训练，企业可以降低成本、掌控自己的数据，并建立差异化的 AI 能力。完整结果发布在其官网链接中。

Current read

AI21 的 CTO 在柏林 GTC 上讲了他们对开源模型做后训练的方法，还拿 Gong、deel 等客户的实际结果做例子，想做私有化部署的可以看看。

11 sources21° AI heatLast updated 2026/10/22 14:47:04

Evidence

11 sources

Primary sourceAI21 Labs

2026/10/22 14:47:04View source

Related source 1IT之家

2026/10/21 10:39:12View source

Related source 2Databricks

2026/10/21 16:00:34View source

Related source 3Artificial Analysis

2026/10/20 15:20:00View source

Related source 4Nathan Lambert

2026/10/20 17:11:08View source

Related source 5Together AI

2026/10/20 20:36:11View source

Related source 6TestingCatalog

2026/10/20 22:02:03View source

Related source 7SuperTechFans

2026/10/21 00:16:33View source

Related source 8Tech in Asia

2026/10/21 05:58:27View source

Related source 9Hugging Face: Blog

2026/10/21 10:59:42View source

Related source 10Merve Noyan

2026/10/22 06:54:10View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story