Isolate intervention timing and causal contribution

Determine how Teacher intervention timing affects collision discovery and whether the Teacher’s guidance causally improves search performance rather than merely correlating with the observed collision outcomes.

Background

The ablation separates Student-only proposals from Teacher-guided proposals and finds that Teacher-guided scenarios account for most discovered collisions. However, this partition does not distinguish the effects of when interventions occur, how often they occur, or whether intervention itself causes the improvement.

A dedicated analysis of intervention timing and causal contribution would clarify the mechanism by which the Teacher affects search behavior.

References

This quantifies the Teacher's aggregate contribution within the reported Student-only and Teacher-guided scenario partition, not intervention timing or diversity evolution, left to future work (Section~\ref{sec:discussion}).

— Teach-to-Crash: A Closed-Loop Student-Teacher LLM Framework for Collision-Inducing Test Scenario Generation  (2609.27296 - Ghazal et al., 23 Sep 2026) in Section 6.1, RQ1: Collision-Finding Effectiveness

The CHR/CDR ablation (Table~\ref{tab:rq1-teacher-ablation}) shows Teacher-guided scenarios account for most collisions within the reported Student-only and Teacher-guided scenario partition; this aggregate association neither shows that intervention causes greater diversity nor measures diversity directly, so we leave a dedicated diversity ablation to future work.

— Teach-to-Crash: A Closed-Loop Student-Teacher LLM Framework for Collision-Inducing Test Scenario Generation  (2609.27296 - Ghazal et al., 23 Sep 2026) in Section 6.3, RQ3: Diversity of Collision Scenarios