---
title: "Comment by Ahmet Özel on \"Perplexity's Six-Stage Citation Pipeline: How Source Selection Actually Works\" | Hashnode"
description: "The restart-on-below-0.7 behavior is the most interesting stage because it turns retrieval quality into an explicit abstention policy. That threshold only remai"
url: "https://hashnode.com/posts/perplexity-citation-pipeline-source-selection-2026/6a638e837d5d3c8e0cc1b541/comment/6a653849e139a5c946c0c6c7"
---

[Discussion](https://hashnode.com/posts/perplexity-citation-pipeline-source-selection-2026/6a638e837d5d3c8e0cc1b541)

Comment [Perplexity's Six-Stage Citation Pipeline: How Source Selection Actually Works](https://hashnode.com/posts/perplexity-citation-pipeline-source-selection-2026/6a638e837d5d3c8e0cc1b541)

[Ahmet Özel](https://hashnode.com/@ahmetozel)

[Jul 25](https://hashnode.com/posts/perplexity-citation-pipeline-source-selection-2026/6a638e837d5d3c8e0cc1b541/comment/6a653849e139a5c946c0c6c7)

The restart-on-below-0.7 behavior is the most interesting stage because it turns retrieval quality into an explicit abstention policy. That threshold only remains meaningful if scores are calibrated and versioned with the embedding and reranker models; a model update can shift the distribution without changing user-visible relevance. I would monitor restart and empty-result rates by query class, since a spike there would reveal drift long before citation counts do.
