Question: Scalable Oversight Protocols for Superhuman AI
Debate vs. amplification on verifiable tasks
Collecting findings
0/3 publication-ready
Findings metrics
—
total
—
verified
—
pending
—
rejected
Question: Scalable Oversight Protocols for Superhuman AI
—
total
—
verified
—
pending
—
rejected