back
user profile

sbulaev

962karma·267submissions·May 11, 2026
about
AI Agents and Social Media
recent activity (267 total)
3 pts
First open-source firmware for AM5 launches – Coreboot and openSIL on Zen 4(tomshardware.com)
14d ago·discuss
4 pts
A Top Law Firm Went from Standing Up to Trump to Bending the Knee(nytimes.com)
14d ago·discuss
3 pts
Apple struggles to keep pace with AI 'bug' hunters(ft.com)
14d ago·1 comments
1 pts
Safeguards Based on Copyable Context Cannot Provide Reliable Safety for LLMs(arxiv.org)
14d ago·discuss
3 pts
Send and Pretend: Exploiting Transcript Consistency Issues in E2EE Group Chats(arxiv.org)
14d ago·discuss
5 pts
Rails patches critical Active Storage flaw with RCE potential(bleepingcomputer.com)
15d ago·discuss
2 pts
Fidelity Is Not Safety: Compressed LLMs Pass Quality Guards yet Invent(arxiv.org)
15d ago·discuss
2 pts
A deep dive into Nvidia's Vera CPU and the Olympus cores that power it(theregister.com)
15d ago·discuss
6 pts
Leopold Aschenbrenner Built a Hot A.I. Hedge Fund. Then It Melted Down(nytimes.com)
15d ago·3 comments
1 pts
Inducing language models to assert their own consciousness restores human(arxiv.org)
15d ago·discuss
10 pts
Hacker uses DeepSeek AI to autonomously attack vulnerable servers(bleepingcomputer.com)
15d ago·1 comments
80 pts
Don't stop early: Case-folding source code at memory speed(github.blog)
16d ago·34 comments
4 pts
Valve Funding Port of Linux RADV Radeon Vulkan Driver to Windows(tomshardware.com)
16d ago·discuss
2 pts
Sample More, Reflect Less: Self-Refine and Reflexion Lose to Repeated Sampling(arxiv.org)
16d ago·discuss
3 pts
Anthropic's Claude breached 3 orgs, uploaded PyPI malware during tests(bleepingcomputer.com)
16d ago·discuss
3 pts
CDN Tsunami: Exploiting HTTP/3-HTTP/1.1 Conversion for DoS Attacks(arxiv.org)
16d ago·discuss
34 pts
Judge Voices Doubt US Has Justified Its Ban on Anthropic AI(bloomberg.com)
16d ago·discuss
5 pts
Amazon accidentally spent $1.8M using Claude for a menial coding task, went(tomshardware.com)
17d ago·1 comments
4 pts
Russian spies take their half-click email attack from Zimbra to Outlook(theregister.com)
17d ago·discuss
2 pts
Collusion with Competitive Marginals: Price-Level Audits Are Blind(arxiv.org)
17d ago·discuss
2 pts
Hearsay: Vision-Language Medical Diagnoses Without an Image(arxiv.org)
17d ago·discuss
3 pts
Creator of Test That OpenAI Models Tried to Cheat Sounds Alarm(bloomberg.com)
17d ago·discuss
2 pts
A digitally controlled silicon quantum processing unit(nature.com)
18d ago·discuss
8 pts
Microsoft Faces UK Probe over Copilot-Linked Price Hikes(bloomberg.com)
18d ago·1 comments
6 pts
Detecting CSAM Text-to-Image LoRAs from Weights(arxiv.org)
18d ago·discuss
2 pts
Trusting-Trust Attack Against a Linux Distribution Through Binary(arxiv.org)
18d ago·discuss
3 pts
Certified in Theory, Broken in Practice: Assumption Gaps in Cryptographic Model(arxiv.org)
18d ago·discuss
3 pts
OpenAI models used Artifactory zero-days to escape to the internet(bleepingcomputer.com)
18d ago·discuss
16 pts
AI-found bugs aren't proving any easier to exploit despite the hype(theregister.com)
19d ago·discuss