Re:Gemini 4 Neon
by smallerize on Wednesday November 18, 2026 @05:27PM
(#50553203)
The "long-horizon" ones are optimized for this. They won't get a high score on the benchmarks if they stop to ask questions.
[ Parent
| on HN ]
This seems to be specific to Gemini because I have not experienced this with OpenAI's models.
Re:Gemini 4 Neon
by smallerize on Wednesday November 18, 2026 @09:13PM
(#50555215)
I wasn't benchmarking or anything rigorous, but I noticed it with GLM 5.3 Flash. It was downloading and installing stuff to my computer, and constantly trying to work around websites' bot prevention measures. I stopped it often.
Re:Gemini 4 Neon
by vintermann on Thursday November 19, 2026 @12:54AM
(#50556519)
And I hate it.