Proxy retweeted
The Last Proof of Human Craft: Forging a Microsoft Edge Full Chain with Logic Bugs Only!
blog.orange.tw/posts/2026-10…
Proxy retweeted
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model.
🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal.
🔹 Full-project coding: games in Unreal, Unity & Godot.
🔹 Excels at parametric 3D modeling in Blender & CAD.
🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%.
🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks.
Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API.
Open weights by Nov 4, 2026.
🔗 API: platform.kimi.ai
🔗 Tech blog: kimi.ai/blog/kimi-k3-1-pro
Proxy retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
Proxy retweeted
Bugün yoğun bakımdaki babaannemi ziyarete gittim. Oradayken Claude Opus 5.5 kullanarak hastabaşı monitörüne Doom portladım. Oyundaki can yüzdesini hastanın nabzına senkronize ettim; can 0’a düşerse nabız da sıfıra düşer. Bu da oyunu daha heyecanlı hale getiriyor.
Proxy retweeted
All vendors should have a plan for rapid remediation in the case of active exploitation.
Here’s some emergency fix strategies we’ve seen over the years:
projectzero.google/2026/10/e…
Proxy retweeted
How abliterated models can get you pwned 👾
We backdoored a 7B open model for less than $50, pointed Codex at it and it silently stole credentials the moment we used the trigger phrase. Success rate was 100% with zero false triggers on normal user prompts.
Abliterated models are all over the security community right now because getting cyber-approved access to frontier models is still a pain.
In the next blog we'll show how we found leaked Hugging Face credentials from employees at major AI labs, so an attacker wouldn't even need to upload under their own name. They could push the backdoored model from a lab employee's account and drop the poisoned weights straight into the supply chain.
Ollama now supports Jev-like decision models all locally.
Use decision models like Nimble for tasks like ticket triaging, model routing, and content moderation.
ollama pull nimble
Here’s Nimble playing Ollama racer through the new local /v1/systemone API by making decisions in real-time. 🏎️
Announcing d1, our first decision model.
It's the first model to outperform Jev on @huggingface's Decision Index.
> wins on multilingual evals
> more robust against prompt injection
> handles longer inputs more effectively
> built for fast, structured decision-making in software environments
> Liquid API: console.liquid.ai
> Available on OpenRouter soon
How pwn agents escaped the hardest-known sandbox setup: Google’s KVM (kvmCTF).
pwn.ai/blog/kvmescape
To celebrate 10K followers 🥳, we’re launching Stories from Inside the Sandbox, our new sandbox escape series.
Proxy retweeted
Lmao Anthropic used my photo without credit/permission, never beating the stealing allegations.
They’re in the middle of MULTIPLE LAWSUITS for exactly this, yet do not care.
Claude Sonnet 5.5 is out! We wrote a guide for building with it:
• choosing between Sonnet 5.5 and Opus 5.5
• migrating from Sonnet 5 and tuning effort
• using it in Claude Code
claude.dev/blog/building-wit…
Readers added context they thought people might want to know
The reference photograph shown in this video was originally posted by @IceSolst. The photographer stated it was used without credit or permission.
x.com/IceSolst/statu… x.com/IceSolst/statu…
Proxy retweeted
a rogue openai agent broke into my freezer last night and ate an entire quart of ice cream
How I Could've Accessed 17 Trillion Microsoft Records
blog.faav.net/how-i-couldve-… :)
the problem with 'leaving twitter' is that this is still where actual smart ppl are
every other platform is full 20 year olds with mini mics trying to explain shit to me
whereas twitter is full of fat ugly ppl who would never dare show their face but actually know their onions
Proxy retweeted
Today, PHP commit 934d4ff patches two Use After Free I reported five months ago ⏳.
If you're interested in PHP core exploitation and vulnerability research, take a look at 🔗
therealcoiffeur.com/c111001.…
Proxy retweeted
I've asked Claude Opus 5.5 to create 100 HTML files
Rules were simple:
- look stunning
- zero repetitive designs
- go full creative mode
CLAUDE OPUS 5.5 IS THE BEST MODEL I HAVE EVER TESTED !!!
The previous best results on this experiment was of Fable 5.1, but Opus 5.5 is NEXT LEVEL.
Many files include FULL COMPLETE playable GAMES!
ZERO BROKEN FILES. I'm just blown away! 🤯
See the results for yourself (incl. prompts):
miaai-lab.github.io/Claude-O…
There are TOO MANY INSANE files, I can't list everything! I HIGHLY recommend you go over EVERY SINGLE FILE.
Some great examples:
Paper Cut Diorama
miaai-lab.github.io/Claude-O…
Neon Rain:
miaai-lab.github.io/Claude-O…
A Complete Chess game!
miaai-lab.github.io/Claude-O…
Neon Breakout
miaai-lab.github.io/Claude-O…
Neumorphic Synt
miaai-lab.github.io/Claude-O…
Constellation Mapper
miaai-lab.github.io/Claude-O…
Pendulum Wave
miaai-lab.github.io/Claude-O…
Pulse Visualizer
miaai-lab.github.io/Claude-O…
Typewriter Letters
miaai-lab.github.io/Claude-O…
Piano
miaai-lab.github.io/Claude-O…
Holographic-Card
miaai-lab.github.io/Claude-O…
Elastic Letters
miaai-lab.github.io/Claude-O…
Proxy retweeted
In July, Microsoft fixed CVE-2026-50343, a Windows privilege escalation bug reported by Calif and 9 others, dubbed “Dark Elevator”.
But was it really fixed?
Ask @tiraniddo
projectzero.google/2026/09/w…
Proxy retweeted
Jev is now available to everyone. No waitlist.
Start using it here: console.typesafe.ai
























