David Gringras

Message

Medical doctor and law graduate. Frank Knox Fellow at Harvard (Health Policy), cross-registered at MIT. Research fellow with Arcadia Impact's AI Governance Taskforce. Currently pivoting from medicine, law, and health policy into AI safety/governance; spending a lot of that time trying to figure out what...

10d

David Gringras hasn't written anything yet.

David Gringras

Message

10d

David Gringras has not written any posts yet.

I Used Safety Benchmarks to Test Agentic Scaffolding. The Benchmarks Were More Broken Than the Scaffolds.

I Used Safety Benchmarks to Test Agentic Scaffolding. The Benchmarks Were More Broken Than the Scaffolds. Epistemic status: Pre-registered, three of five directional predictions disconfirmed, failures published. Format dependence started exploratory and was validated through a separately pre-registered confirmatory experiment (18 falsification tests, zero failures). All code, data, and pre-registrations...

Feb 2410

David Gringras hasn't written anything yet.

LESSWRONG
LW

LESSWRONG
LW

David Gringras

David Gringras

David Gringras

David Gringras

I Used Safety Benchmarks to Test Agentic Scaffolding. The Benchmarks Were More Broken Than the Scaffolds.

I Used Safety Benchmarks to Test Agentic Scaffolding. The Benchmarks Were More Broken Than the Scaffolds.