#neural controlled differential equations

topicneural controlled differential equations

1 paper · 1 filter

cs.AI2026

Tracing Agentic Failure from the Flow of Success

Samuel Yeh, Yiwen Zhu, Shaleen Deep +1

The paper introduces OAT, a lightweight unsupervised method that learns from successful LLM agent trajectories and detects error steps in failed runs by scoring deviations using ne…