mlsanityData to 14 Oct 2026

Topic Is Not Agenda: A Citation-Community Audit of Text Embeddings

arXiv 2605.07158 ↗First posted 8 May 2026cs.IR cs.CL cs.LGcited by 0

On the map: Retrieval-augmented generation → Knowledge-graph and structured-knowledge RAG

Abstract (arXiv)

Vector search and retrieval-augmented generation (RAG) rest on the assumption that cosine similarity between text embeddings reflects conceptual relatedness. We measure where this assumption breaks. We build an augmented citation graph over 3.58M scientific papers and partition it via Leiden CPM at two granularities: sub-field (L1) and research-agenda (L2, hierarchical inside each L1). Four state-of-the-art embeddings (Gemini, Qwen3-8B, Qwen3-0.6B, SPECTER2) clear the L1 bar reasonably (45-52% top-10 same-rate) but stop working at L2: only 15-21% of top-10 neighbors share the query's research agenda. In absolute terms, 8 of every 10 retrieved papers are off-agenda. The failure is universal across eight scientific domains and all four models; SPECTER2, despite its citation-based contrastive training, is the weakest. As a diagnostic probe, we test whether the same augmented graph also functions as a retrieval signal: a deliberately simple citation-count rerank reaches 57.7% top-1 L2 on top of LLM-expanded Boolean retrieval and 59.6% on top of plain BM25, on 80 curated agenda queries -- about 9 points above the best cosine retriever (Gemini, 50.6%) and 20 points above BM25 alone (39.3%). The probe isolates a slice of the agenda-matching signal the graph carries but the embeddings miss, connecting recent theoretical limits on single-vector retrieval to a concrete failure mode of scientific RAG.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Knowledge-graph and structured-knowledge RAG, one of 15 directions in Retrieval-augmented generation. Builds on 12 of its 15 references that are in the map. First posted on arXiv 8 May 2026.

Builds on

12 of 15 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024From Local to Global: A Graph RAG Approach to Query-Focused Summarization24 Apr

Other references · 11

  1. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  2. 2017Billion-scale similarity search with GPUscited by 2,497
  3. 2022MTEB: Massive Text Embedding Benchmarkcited by 640
  4. 2022Precise Zero-Shot Dense Retrieval without Relevance Labelscited by 517
  5. 2020SPECTER: Document-level Representation Learning using Citation-informed Transformerscited by 430
  6. 2018From Louvain to Leiden: guaranteeing well-connected communitiescited by 385
  7. 2022SciRepEval: A Multi-Format Benchmark for Scientific Document Representationscited by 129
  8. 2025On the Theoretical Limitations of Embedding-Based Retrievalcited by 104
  9. 2025Gemini Embedding: Generalizable Embeddings from Geminicited by 94
  10. 2024Is Cosine-Similarity of Embeddings Really About Similarity?cited by 89
  11. 2022Neighborhood Contrastive Learning for Scientific Document Representations with Citation Embeddingscited by 75

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Citegeist: Automated Generation of Related Work Analysis on the arXiv Corpus
  2. 2023ATLANTIC: Structure-Aware Retrieval-Augmented Language Model for Interdisciplinary Science
  3. 2026Graph-Aware Late Chunking for Retrieval-Augmented Generation in Biomedical Literature
  4. 2025From Papers to Panoramas: Building Hierarchies of Scientific Literature at Scale
  5. 2025CG-RAG: Research Question Answering by Citation Graph Retrieval-Augmented LLMs
  6. 2026From Association to Causation: Improving Retrieval Precision of Retrieval-Augmented Generation via Causal Relations and an Attention Mechanism
  7. 2024Beyond Benchmarks: Evaluating Embedding Model Similarity for Retrieval Augmented Generation Systems
  8. 2026Does Document Structure Help Dense Retrieval? A Placebo-Controlled Ablation of Four Mechanisms Across Two Corpora