renv
renv/span-citationpublicv1 · Jul 20, 2026e4f1a9c

Span-Anchored Citations for Verifiable Research Writing

Citations that point at papers are unverifiable in practice: the reader cannot find the sentence that supports the claim. We anchor every citation to the exact quoted source span (W3C TextQuoteSelector), verify that the span supports the claim, and emit self-verifying LaTeX citations. The anchor is the quoted text itself, so citations survive re-indexing and new document versions.

Interactive graphRepositorycitationsverifiabilityretrievalNLP
3/4

claims supported

3 · 3

experiments · runs

3

span citations

3

referenced papers

Claims & evidence

Status is derived from linked evidence, never asserted.

thesis

Anchoring citations at the span level makes them self-verifying: the quoted text either exists in the source or it does not.

§1supported
  • ⊕ supports003-anchor-robustnesslocalanchor_survival_rate 0.97confirmatory
  • ⊕ supportsZhang 2024 · arXivfull · 0.90
    fine-grained sentence-level citations allow users to locate the exact supporting evidence
contribution

Sentence-granularity passages are easier to verify than paragraph chunks without losing retrieval recall.

§4.2supported
  • ⊕ supports001-tfidflocalrecall_at_8 0.81span_support_rate 0.74confirmatory
  • ⊕ supportsGao 2023 · EMNLPfull · 0.92
    smaller retrieval units are easier for humans to verify against the generated statement
  • ⊕ supportsLei 2024 · arXivpartial · 0.63
    verification accuracy degrades as the evidence context grows
contribution

Span anchors survive re-indexing and new document versions because they never reference chunk or vector ids.

§3.3supported
  • ⊕ supports003-anchor-robustnesslocalanchor_survival_rate 0.97confirmatory
claim

Dense retrieval is required for competitive recall on claim-to-source matching.

§4.1refuted
  • ⊗ refutes002-denselocalrecall_at_8 0.84span_support_rate 0.76suggestiveDense beats TF-IDF by only 3 points recall@8, so it is not strictly required.

Experiments & results

Every number traces to a recorded run with git sha, env hash, and seed.

ExperimentHypothesisRunMetrics
001-tfidf
TF-IDF baseline
A stdlib TF-IDF retriever reaches usable recall@8 on claim→span matching.
b31c2ffseed 7local
recall_at_8 0.81 (test)
span_support_rate 0.74 (test)
002-dense
Dense embeddings001-tfidf
Dense retrieval (BGE-M3) beats TF-IDF by a wide margin.
d02e881seed 7local
recall_at_8 0.84 (test)
span_support_rate 0.76 (test)
003-anchor-robustness
Anchor robustness across versions001-tfidf
Quote-text anchors re-resolve after re-parsing with a different PDF extractor.
e4f1a9cseed 7local
anchor_survival_rate 0.97 (test)

Papers & span citations

Citations anchor to the exact quoted span, so every quote can be checked against its source.

Enabling Large Language Models to Generate Text with CitationsTianyu Gao, Howard Yen, Jiatong Yu et al. · 2023 · EMNLP
gao2023alce

We observe that smaller retrieval units are easier for humans to verify against the generated statement, motivating sentence-level attribution.

chars 14210–14298 · support: full · 0.92

LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QAJiajie Zhang, Yushi Bai, et al. · 2024 · arXiv
zhang2024longcite

In contrast to passage-level attribution, fine-grained sentence-level citations allow users to locate the exact supporting evidence within long documents.

chars 3120–3211 · support: full · 0.90

FactCG: Enhancing Fact Checkers with Graph-based Multi-hop DataDeren Lei, Yaxi Li, et al. · 2024 · arXiv
factcg2024

verification accuracy degrades as the evidence context grows

support: partial · 0.63

Code references

Deep links to the exact lines, pinned to the published commit.