Event ·Multi-source

Harvey后训练Nemotron 3.5 Lightning，法律Agent基准得分8.3%

Harvey联合Trajectory Labs对NVIDIA的Nemotron 3.5 Lightning进行后训练，在Legal Agent Bench上得分从0%升至8.3%。该成绩超过了Opus 4.6和更大的后训练版Nemotron 3 Ultra。后训练覆盖九个执业领域，未出现能力退化。平均输出长度从90k tokens降至37k tokens，每token奖励提升2.4倍。

Current read

Harvey把Nemotron 3.5 Lightning后训练了一把，法律Agent评测从0冲到8.3%，比Opus 4.6还强，输出也从90k缩到37k，省Token。

11 sources71° AI heatLast updated 2026/8/12 21:47:16

Evidence

11 sources

Primary sourceNVIDIA AI

2026/8/12 21:47:16View source

Related source 1ollama

2026/8/11 16:02:21View source

Related source 2Aravind Srinivas

2026/8/12 01:37:26View source

Related source 3Fireworks AI

2026/8/11 14:49:20View source

Related source 4Decoder

2026/8/11 15:07:44View source

Related source 5AI Will

2026/8/12 06:41:07View source

Related source 6OpenRouter

2026/8/11 13:17:30View source

Related source 7Harrison Chase

2026/8/11 17:23:48View source

Related source 8IT之家

2026/8/12 01:49:32View source

Related source 9marktechpost

2026/8/12 06:59:40View source

Related source 10Julien Chaumond

2026/8/12 16:16:12View 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