1Cademy - LectureBank-Full Target-Disjoint R@10 Result (n=164): Diffusion Gain Survives, Adaptive Tied

Learn Before

LectureBank-Full Prerequisite QA Benchmark
Method Part 2: Question-Disjoint and Target-Concept-Disjoint (Auditable Strict-Parity Graph-RAG Paper)
Method Part 2: Bidirectional Prerequisite Diffusion with Role-Aware (Auditable Strict-Parity Graph-RAG Paper)
Leakage Controls (Results) in Auditable Strict-Parity Evaluation of Prerequisite-Graph Retrieval for RAG under Leakage Controls

Example

LectureBank-Full Target-Disjoint R@10 Result (n=164): Diffusion Gain Survives, Adaptive Tied

Under the target-concept-disjoint control on LectureBank-Full ( $n = 164$ ), R@ $10$ is $56.9\%$ $[53.9, 60.1]$ for flat dense, $88.1\%$ $[84.9, 91.2]$ for the fixed-depth hierarchical baseline, and $89.9\%$ $[86.7, 93.0]$ for Adaptive (heuristic). The paired-bootstrap delta of Adaptive $-$ Hierarchical at $k=10$ is $+1.8$ $[-0.9, +4.7]$ . The large flat-to-hierarchical gain attributable to bidirectional diffusion with role-aware quotas survives the stricter split, while Adaptive remains statistically tied to the fixed-depth baseline.

Updated 2026-05-17

Contributors are:

Who are from:

References

Reference: Auditable Strict-Parity Evaluation of Prerequisite-Graph Retrieval for RAG under Leakage Controls

Learn Before

Related