Agent prompt for the lora task: > From first principles, in a new repo that can reference the compvis one, make an efficient lora trainer for SD 1.5 that works on a cpu. This is the original lora paper: https://
arxiv.org/html/2106.0968
5v2
…, this is the Stable Diffusion paper:
AI
-
Efficient CPU LoRA Trainer for SD 1.5 from First Principles
By
–
-
Efficient CPU LoRA Trainer for SD 1.5 based on CompVis repo
By
–
Agent prompt for the lora task: > From first principles, in a new repo that can reference the compvis one, make an efficient lora trainer for SD 1.5 that works on a cpu. This is the original lora paper: https://
arxiv.org/html/2106.0968
5v2
…, this is the Stable Diffusion paper: -
New paper and code on sequential poisoning attacks in AI
By
–
Lots more in the paper: how does DPO fit into the picture? What if attackers have different goals? etc. Paper: https://
arxiv.org/abs/2606.04929
Code: https://
github.com/jcksanderson/s
equential-poisoning
… Led by @jcksanderson
, w/ @YihanWww
, Xiaoqian Lu, co-supervised w/ @YiweiLu3r 6/6 -
0.5% poison breaks reward model, 5% needed for RLHF transfer
By
–
What about poisoning PPO? A remarkable paper of @javirandor and @florian_tramer (
https://
arxiv.org/abs/2311.14455) shows that just 0.5% poison is enough to break a reward model (L)! Again, fear not: somehow, it takes a (high) 5% poisoning before it transfers to the RLHF'd model (R). 4/n -
2% SFT poisoning gives 90% attack success; RLHF wipes it away
By
–
There's multiple post-training phases attackers can infiltrate: SFT, DPO, PPO. Let's start with SFT. With just 2% SFT poisoning, 90% attack success (L)! But not to worry, RLHF works as we hope (?): it wipes away the poison. An RM scores outputs just like a clean model (R). 3/n
-
LLM post-training pipelines vulnerable to combined data poisoning attacks
By
–
Feeling safe against data poisoning in post-training? Think again! Individual components of LLM post-training pipelines are surprisingly robust to data poisoning attacks. In work led by @jcksanderson (co-advised w @YiweiLu3r
), we show they crumble when attacked together. 1/n -
Backdoor attacks on LLMs via untrusted training data
By
–
LLMs are trained on lots of data, often from untrusted sources. This is particularly true in safety post-training, where data is gathered from human responses. Attackers can try to sneak in a backdoor: if there's a trigger in the prompt, bypass safety guardrails. 2/n
-
Free 2-hour masterclass: build entire startup using Claude Design
By
–
THIS GUY IS LITERALLY GIVING AWAY THE DESIGN PLAYBOOK FOR CLAUDE DESIGN 🤯
— Charly Wargnier (@DataChaz) 5 juin 2026
A Free 2-hour masterclass showing how to build an ENTIRE startup:
→ brand guidelines
→ decks
→ website
→ apps
→ videos
.. using ONLY Claude Design.
full 2 hour tutorial + guide below in 🧵 ↓ https://t.co/adr7Pa4jjs pic.twitter.com/pwwDTJCkesTHIS GUY IS LITERALLY GIVING AWAY THE DESIGN PLAYBOOK FOR CLAUDE DESIGN A Free 2-hour masterclass showing how to build an ENTIRE startup: → brand guidelines
→ decks
→ website
→ apps
→ videos .. using ONLY Claude Design. full 2 hour tutorial + guide below in ↓ -
Testing Gemini 3.5 Flash and Antigravity CLI with Stable Diffusion 1.5
By
–
Today I'm experimenting with Gemini 3.5 Flash and the Antigravity CLI to see how fast and how autonomously the agents can do things. – It took 20 minutes to install and run the original CompVis Stable Diffusion 1.5 repo, get the weights, debug, run inference and generate an
-
AI Agents Shift Workflows from Tools to Coordinated Human-Agent Work
By
–
AI Agents Are Coming For Your Workflows. I’ve spent the day at Asana’s Work Innovation Summit in London, where the big theme was the next phase of enterprise AI: moving from individual productivity tools to coordinated human-agent work. #Sponsored
— Bernard Marr (@BernardMarr) 5 juin 2026
The key message for me was… pic.twitter.com/uS7Ki8oIzPAI Agents Are Coming For Your Workflows. I’ve spent the day at Asana’s Work Innovation Summit in London, where the big theme was the next phase of enterprise AI: moving from individual productivity tools to coordinated human-agent work. #Sponsored The key message for me was