For anyone who was piqued by the title but couldn't stomach the unedited LLM blogpost, you can read the similarity AI-genned docs for Laya directly: https://github.com/NandhaKishorM/laya
Similar result in production for me. I compared Jev with a general LLM (DeepSeek) as the decider for when to cut live speech into sentences, on the same 226 segments. With Jev the longest held segment was 17 s. With DeepSeek it was 40 s.
The flood of posts related to this thing does not seem organic.
For anyone who was piqued by the title but couldn't stomach the unedited LLM blogpost, you can read the similarity AI-genned docs for Laya directly: https://github.com/NandhaKishorM/laya
Man, it shows that someone was getting paid by the word.
Thank you!
Similar result in production for me. I compared Jev with a general LLM (DeepSeek) as the decider for when to cut live speech into sentences, on the same 226 segments. With Jev the longest held segment was 17 s. With DeepSeek it was 40 s.
I tried both with a synthetic dataset (1000 items), Jev scored 98% vs Laya 15%.
[flagged]
[flagged]