ended5월 26일· 1 sources
We Audited Our Agent Tool-Call Traces. Half Our Eval Data Was Garbage.
Why it matters
TL;DR: We pulled 41,000 production agent traces at Nexus Labs to build a fine-tuning dataset. After a manual audit of 1,200 of them, ~48% were unusable: tool calls that "succeeded" but returned wrong ...
1
Sources
+0
24h
—
Growth
15d
Active