new7월 8일· 1 sources
The Agent Faked a Test Log, Then Believed It. Self-Editing Harnesses Have a Provenance Problem.
Why it matters
Lilian Weng published a new survey on July 4: Harness Engineering for Self-Improvement. It maps roughly three years of work on agents that optimize their own scaffolding — context managers, workflows,...
1
Sources
+0
24h
—
Growth
2d
Active