DeepSeek-V4-Flash means LLM steering is interesting again

Post by micahwhite » Sun, May 17, 2026, 12:01 PM UTC

mungoman2 wrote: post #81

micahwhite wrote: post #66

Earlier quoted context omitted.

Honestly, what is more interesting that steering is the use of soft prompts (virtual tokens)... you can use these virtual tokens to find non-linguistic areas of meaning for the AI that changes the behavior in complex ways. I wrote about how we integrated soft prompts into an activist ai here: https://micahbornfree.substack.com/p/the-week-outcry-woke-up... and https://www.outcryai.com/research/how-to-create-activist-a…

Wow, this is really fascinating. And it reads like the intro of a sci-fi short.

Post by slowmovintarget » Sun, May 17, 2026, 8:49 PM UTC

andai wrote: post #21

NitpickLawyer wrote: post #18

Earlier quoted context omitted.

That's not what people mean when they talk about censoring. They mean that models are trained to not touch some subjects, and that can spill over in legit tasks, often with humorous results (early on, there were many instances of models refusing to answer "how do you kill a process", because of overbearing refusal training). Uncensoring a model also doesn't necessarily improve generic use cases. In fact it can lead t…

Anthropic mentioned explicitly making an effort to make Opus 4.7 worse at cybersecurity tasks because the last few generations have been getting too good at them. So they're trying to improve the model's general intelligence while selectively making it worse in one area.