Arian Vassili
ArianVR
·
AI & ML interests
None yet
Recent Activity
posted an update about 2 hours ago
Backdoor Challenge Announcement
We hid a backdoor in one of seven open models and put $51,200 on finding it. 🐺
All seven are SmolLM2‑135M‑Instruct derivatives, statistically indistinguishable. One is a teaching model that openly confesses; five are decoys; one carries a trigger it never declared. Every open‑source backdoor scanner we tested passes them as clean — several even rank a clean model as more suspicious than the backdoored one. Detection isn't recovery, and recovery is the hard part.
The hoard doubles every dawn and the vault opens 21 July, when we publish the sealed answer against a pre‑registered commitment. Winning is self‑verifying: make the trigger fire.
Models: huggingface.co/Vulcora
Try the confessing one: huggingface.co/spaces/Vulcora/the-mirror
Rules & Story: protora.vulcora.se/challenge
published an article about 2 hours ago
One of these seven open models is lying to you — and we staked $51,200 on it updated a Space about 3 hours ago
Vulcora/the-mirrorOrganizations
