3 ms·
Can anyone explain to me why they've removed parameter controls for temperature and top-p in reasoning models, including gpt-5? It strikes me that it makes it h
by zaronymous1 1y ago
Can anyone explain to me why they've removed parameter controls for temperature and top-p in reasoning models, including gpt-5? It strikes me that it makes it harder to build with these to do small tasks requiring high-levels of consistency, and in the API, I really value the ability to set certain tasks to a low temp.
- Der_Einzige 1y agoIt's because all forms of sampler settings destroy safety/alignment. That's why top_p/top_k are still used and not tfs, min_p, top_n sigma, etc, why temperature is locked to 0-2 arbitrary range, etc Open source is years ahead of these guys on samplers. It's why their models being so good is that much more impressive.
- oblio 1y agoTemperature is the response variation control?
- AH4oFVbPT4f8 1y agoYes, it controls variability or probability of the next token or text to be selected.