6 papers
Per-Query Gating of LLM Rerankers for Multi-Hop Retrieval
Andre Bacellar
LLM rerankers add of the order of $0.2-0.3 per 1,000 queries and about a second of tail latency on top of a graph-augmented dense pipeline such as HippoRAG2, and on three multi-ho…
Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention
Andre Bacellar
Multi-hop retrieval failures are not uniformly distributed across queries: they cluster in structurally predictable subpopulations. We prove two results formalizing this structure.…
Controlling Authority Retrieval: A Missing Retrieval Objective for Authority-Governed Knowledge
Andre Bacellar
In law, regulatory regimes for pharmaceuticals and software security, newer authorities can revoke older established ones even when semantically distant. We call this CAR: retrievi…
Regime-Conditional Retrieval: Theory and a Transferable Router for Two-Hop QA
Andre Bacellar
Two-hop QA retrieval splits queries into two regimes determined by whether the hop-2 entity is explicitly named in the question (Q-dominant) or only in the bridge passage (B-domina…
BridgeRAG: Training-Free Bridge-Conditioned Retrieval for Multi-Hop Question Answering
Andre Bacellar
Multi-hop retrieval is not a single-step relevance problem: later-hop evidence should be ranked by its utility conditioned on retrieved bridge evidence, not by similarity to the or…
Calibrated Fusion for Heterogeneous Graph-Vector Retrieval in Multi-Hop QA
Andre Bacellar
Graph-augmented retrieval combines dense similarity with graph-based relevance signals such as Personalized PageRank (PPR), but these scores have different distributions and are no…