Re:Ollaya – Ollama for open-source, Jev-style decision models
by OgAstorga on Friday September 25, 2026 @03:15PM
(#49848695)
text classification is equivalente to decision. This is exactly the same thing Jev does.
[ Parent
| on HN ]
Re:Ollaya – Ollama for open-source, Jev-style decision models
by ranyume on Friday September 25, 2026 @03:22PM
(#49848776)
If it has four legs, a tail and barks why not call it a dog?
Re:Ollaya – Ollama for open-source, Jev-style decision models
by gchamonlive on Friday September 25, 2026 @03:30PM
(#49848877)
Because this specific dog only barks in structured text
Re:Ollaya – Ollama for open-source, Jev-style decision models
by seemaze on Friday September 25, 2026 @04:20PM
(#49849410)
This dog only barks when given biscuits
Re:Ollaya – Ollama for open-source, Jev-style decision models
by ricardobeat on Friday September 25, 2026 @03:21PM
(#49848769)
It is not. In a benchmark with actual decisions - navigation, traffic, waypoints - laya does slightly better than a small classifier, with very low correlation to state changes.
Re:Ollaya – Ollama for open-source, Jev-style decision models
by abirch on Friday September 25, 2026 @03:18PM
(#49848741)
Jev does it more efficiently because it doesn't use an LLM
https://typesafe.ai/blog/introducing-system-one-models-and-j...
Re:Ollaya – Ollama for open-source, Jev-style decision models
by rockinghigh on Friday September 25, 2026 @04:12PM
(#49849305)
Their marketing language is misleading. They must still use some transformer language model backbone to encode the text input (BERT or decoder-only LLM). The biggest difference is the output, instead of auto-regressively generating tokens, they produce probabilities over a bounded set of decisions (more flexible classification).