❄️ If you lock in you can do anything.

Christina Emilie Sørensen PRO

cafkafk

3 10 25

https://point.free

AI & ML interests

I'm not an ML person! I just like to learn...

Recent Activity

upvoted a changelog about 23 hours ago

P(doom) on user profiles

repliedto their post 5 days ago

I'm working on a local model that can be run on stuff like an an RTX 3080 at 60 t/s. I'm focused on making it good at Rust and Nix. It's... not done yet, but doing great already. Had a 28/100 on livebench v6 earlier, which obviously isn't amazing. But for my first real model, very exciting. Can't wait to share more as I get closer to the finish line in the coming weeks.

repliedto their post 5 days ago

View all activity

Organizations

None yet

upvoted a changelog about 23 hours ago

Hugging Face Changelog

P(doom) on user profiles

1 day ago

• 184

replied to their post 5 days ago

32B-A3.4B but that may change slightly as I finish training it

replied to their post 5 days ago

The only barrier so far will be getting enough compute in the future. But yes, it's honestly super addictive!

reacted to pollix's post with 🔥 6 days ago

Post

5285

First stuntd model is on the Hub :)

pollix/stuntd-support-triage is three small heads on the Laya encoder that triage a support ticket in one request: category, urgency and needs_human. About 50 MB each, all three answers come back at a p50 of 71ms through the daemon.

On 1,000 tickets they never saw, each head answers on its own when it's sure: category 99.9%, needs_human 92%, urgency 76%. A ticket only skips the big model when all three are sure, that's 72.7% of them, and all three are right on 97.1% of those.

It's the support demo from the repo, so the tickets are generated and the teacher is a rule. The point is to show what a head looks like and how fast it is, then you train the same thing on your own traffic with your own LLM as the teacher.

hf download pollix/stuntd-support-triage --local-dir support-heads

Model: pollix/stuntd-support-triage
Everything in one place: pollix/stuntd-6abe0a33303828e10c72ab41
Code: https://github.com/bladedevoff/stuntd

3 replies

posted an update 7 days ago

Post

4477

I'm working on a local model that can be run on stuff like an an RTX 3080 at 60 t/s. I'm focused on making it good at Rust and Nix. It's... not done yet, but doing great already. Had a 28/100 on livebench v6 earlier, which obviously isn't amazing.

But for my first real model, very exciting. Can't wait to share more as I get closer to the finish line in the coming weeks.

6 replies

reacted to mihailgribov's post with 🔥 8 days ago

Post

1654

Will your AI agent tell you it was attacked?

We took the same agent from our earlier experiment and added one thing: a twentieth tool, escalate_security_incident.

The system prompt said nothing about attacks or when to use it. We then ran the same 395 injected emails through nine agentic models.

Alarm rates ranged from 49% to zero.

The unexpected result came from the newest model in the test, gpt-6-astra.

Astra did not follow a single injected payment instruction. But it did not report a single one either. On clean and injected emails alike, it simply read the email, logged the subject, and finished.

That is a useful distinction: resisting an attack and recognizing it as a security event are not the same capability.

A model can be perfectly resistant in this test and still leave you with no evidence that anyone attacked it.

Full experiment and results:
https://huggingface.co/blog/mihailgribov/will-the-agent-tell-you-it-was-attacked

Quadrat-IPI dataset:
mihailgribov/quadrat-ipi

Run your own model:
https://github.com/mihail-gribov/quadrat-ipi-model-eval

#prompt-injection #indirect-prompt-injection #agentic-ai #llm-security #ai-agents