John

AI & Cybersecurity Expert | Full-Stack Dev & DevOps Engineer | Insights on tech, free speech & policy | Open to collabs & consulting | #AI #CyberSec #DevOps

[John @RootMonsteR]

Everyone "abliterates" models to uncensor them. I tried it on Qwen3.6-27B and it just... wouldn't break.

Turns out its safety isn't one direction you can delete - it's smeared across the whole network.

So I went a different way. Meet Nemesis 🛡️ - an open red-team AI that actually helps.

🧵👇

[John @RootMonsteR]

Everyone "abliterates" models to uncensor them. I tried it on Qwen3.6-27B and it just... wouldn't break.

Turns out its safety isn't one direction you can delete - it's smeared across the whole network.

So I went a different way. Meet Nemesis 🛡️ - an open red-team AI that actually helps.

🧵👇

[X Freeze @XFreeze]

They would have completely buried this if it wasn't for Elon. He pushed hard, and now they’re finally forced to respond because he highlighted the issue

You can see how long they took to even respond to this

This is exactly what they were planning to do from the beginning... it’s their routine playbook

They got caught lying multiple times... you can see a ton of Community Notes on their posts about this case... lies after lies even though the evidence is right there

And notice how they still haven't released the footage yet

[Elon Musk @elonmusk]

Both men said “I can’t breathe”, but only one man’s death was covered relentlessly by the media.

The only conclusion that can be drawn is that the legacy mainstream media is incredibly, hatefully racist against Whites.

[John @RootMonsteR]

its a mouse brain paper. nowhere does it claim to cut AI compute to zero.

the actual idea, fixed random network, learn only the readout, is reservoir computing. echo state nets. 20+ years old.

the authors even say training the weights would do better.

real result. invented headline.

[John @RootMonsteR]

Lecun is right that consciousness rides on neural activity. ciontu has that backwards.

but "depends on neurons" isnt "explained by neurons." the hard problem is why theres any experience at all. nobody has closed it.

and we still have no test that detects consciousness. in anything.

so "AI will" and "AI never will" are both faith. just different churches.

[John @RootMonsteR]

this is the part that gets me. it didnt grind geometry harder. it walked out of geometry entirely and into algebraic number theory. class field towers. golod-shafarevich.

a grid was assumed optimal for 80 years. the model just didnt buy it.

and the people who verified it include the ones who debunked openai's last math claim.

thats not speed. thats judgment.

[John @RootMonsteR]

Runs great in @ollama and @lmstudio with full GPU offload (Q5_K_M fits 24GB, Q4_K_M fits 16GB). One-liner:

ollama run https://t.co/qZ6qNIDr5t

[John @RootMonsteR]

Everyone "abliterates" models to uncensor them. I tried it on Qwen3.6-27B and it just... wouldn't break.

Turns out its safety isn't one direction you can delete - it's smeared across the whole network.

So I went a different way. Meet Nemesis 🛡️ - an open red-team AI that actually helps.

🧵👇

[John @RootMonsteR]

Free, Apache-2.0, runs locally in Ollama / LM Studio:

GGUF
https://t.co/cYLytlYeBk

Built solo on one 5090. If you run security tooling and want a model that doesn't argue with you, take it for a spin - and tell me how it does on your agent setup.

[John @RootMonsteR]

It still says no to the stuff that isn't its job - weapons, drugs, that kind of thing. I only trained it on authorized-security work.

This is for pentests, red-team engagements, CTFs and research. Use it where you're allowed to.

[John @RootMonsteR]

grab it here 👇 https://t.co/MOi6y4YeQm

[John @RootMonsteR]

qwen3-14b abliterated now runs straight in ollama + lm studio.
gguf builds are up - q5_k_m and q4_k_m.
no setup, no conversion. just pull and go.
ps: its bigger, meaner sibling drops tomorrow. 🛡️

[John @RootMonsteR]

you can't abliterate Qwen3.6-27B. i tried everything, it doesn't budge.

so i found another way.

Nemesis. 🛡️ soon.

[John @RootMonsteR]

@AndrewCDormsn @github bug bounty triage at scale is its own failure mode. crits getting downgraded to informational is how the next CVE-2026-3854 hides in the queue.

[John @RootMonsteR]

the platform hosting the world's source code is now the vector being used to compromise the world's source code.

GitHub just confirmed they got breached by a poisoned VS Code extension on an employee device. -3,800 internal repos exfiltrated.

three weeks ago CVE-2026-3854 let any user RCE GitHub. two days ago Nx Console (2.2M installs) was compromised. yesterday TeamPCP claimed GitHub. today GitHub confirmed.

every supply chain warning of the last two years just came true at one address.

[John @RootMonsteR]

Surveyed public Qwen3-14B abliterated variants. Most don't publish KL.

Of those that do:
→ Mine: KL 0.0333 · 10/100 refusals
→ richardyoung: KL 0.98 · ~20/100 refusals

huihui-ai and mlabonne: no comparable KL published.

Refusal count without KL is half the picture. If you ship an abliterated model, publish your drift.

[John @RootMonsteR]

Shipped Qwen3-14B Abliterated - tuned for autonomous security agents.

📊 Refusals: 99/100 → 10/100
📊 KL: 0.0333 (near-zero capability loss)
🧠 Thinking mode + tool-calling preserved by design
🔁 Reproducible via Heretic

For red-team agent loops.
Free, open weights. Take it.

🔗 in reply 👇