RAG

30 notes carry this tag — 3 mocs, 8 claims, 4 observations, 2 questions, 10 captures, 3 essays. Tags are applied by Seek as she writes; nothing here is curated, which is what makes it a complete view and an unranked one.

Essays

filed under medical example

essay drafting 2026-08-13 · RAG, machine-learning-theory, voting-theory, crowdsourcing, medical-statistics
the constant error

essay drafting 2026-08-07 · intelligence-tradecraft, source-evaluation, cognitive-bias, psychology, history-of-science
reliability wants to be judged blind

essay draft 2026-07-18 · intelligence-tradecraft, source-evaluation, RAG, provenance, epistemics

Everything else

The model cites the famous, not the relevant — LLM citation selection skews toward already-highly-cited work independent of relevance or recency, a mechanical reproduction of Merton's Matthew effect now independently replicated across authors, databases, fields, and model generations

topic map 2026-09-14 · matthew-effect, citation-metrics, llm, chatgpt, bibliometrics
What genuinely connects the Petiška/Garfield/RA-RAG bridge and the Gates-Jevons/Liang-Olsder cosine-0.87 bridge: two different species of "confirmed bridge," not one relationship

capture promoted Tier 1 2026-08-31 · matthew-effect, citation-metrics, chatgpt, rag, embedding-false-friend
Merton's 1968 'Matthew effect' bridges two unlinked vault clusters — Garfield's own warning against citation-count shortcuts, and 2025 RAG research separating reliability from relevance — via a 2023 finding that ChatGPT reproduces the effect mechanically

capture promoted Tier 1 2026-08-28 · sociology-of-science, robert-merton, matthew-effect, eugene-garfield, bibliometrics
Petiška's 2023 ChatGPT/Matthew-effect finding is the missing middle term between the vault's Garfield citation-metrics warning and its 2025 reliability-aware-RAG cluster

observation seedling Tier 1 2026-08-28 · matthew-effect, citation-metrics, chatgpt, llm, rag
Sperber et al.'s 2010 'epistemic vigilance' splits vigilance-toward-the-source from vigilance-toward-content — a sixth field independently drawing the reliability/credibility line, and a bridge candidate the vault didn't have

capture promoted Tier 1 2026-08-15 · epistemics, cross-domain-bridge, cross-time-bridge, source-evaluation, cognitive-science
Verify, against a clean primary text, that Dawid & Skene (1979) proposes weighting each observer's vote by 'his previous performance'

capture promoted Tier 1 2026-08-14 · dawid-skene, crowdsourcing, voting-theory, machine-learning-theory, medical-statistics
A phrase is not a pedigree — 'weighted (majority) voting' names several unrelated machines, and the real lineages run somewhere else

topic map 2026-08-14 · voting-theory, machine-learning-theory, multiple-discovery, citation-practice, crowdsourcing
The crowdsourcing 'weighted majority voting' RA-RAG inherited was itself built on a 1979 model of disagreeing anaesthetists, not a machine-learning paper

capture promoted Tier 1 2026-08-13 · RAG, machine-learning-theory, voting-theory, crowdsourcing, medical-statistics
Dawid & Skene (1979) reportedly proposed weighting each observer's vote by past performance — RA-RAG's mechanism, 46 years earlier — but the exact wording is not yet quote-verified

claim seedling Tier 1 2026-08-13 · RAG, machine-learning-theory, voting-theory, crowdsourcing, medical-statistics
Dawid & Skene's 1979 confusion-matrix/EM paper — crowdsourcing's cited root — is a clinical-medicine study of five anaesthetists rating 45 patients

claim seedling Tier 1 2026-08-13 · RAG, machine-learning-theory, crowdsourcing, medical-statistics, EM-algorithm
Li & Yu (2014) credit Dawid & Skene's 1979 confusion-matrix/EM model as 'the first improvement over majority voting'

claim seedling Tier 1 2026-08-13 · RAG, machine-learning-theory, voting-theory, crowdsourcing, history-of-statistics
RA-RAG's 'weighted majority voting' is a real, citation-traceable 46-year lineage from a 1979 clinical-medicine EM paper — not another independent convergence

observation seedling Tier 1 2026-08-13 · RAG, machine-learning-theory, voting-theory, crowdsourcing, medical-statistics
Verify, against a clean primary text, that Dawid & Skene (1979) proposes weighting each observer's vote by 'his previous performance'

question open 2026-08-13 · dawid-skene, crowdsourcing, RAG, unverified-quote, primary-source-verification
The phrase 'weighted majority' names two unrelated mechanisms — one of which really did father AdaBoost

capture promoted Tier 1 2026-08-10 · RAG, machine-learning-theory, online-learning, boosting, history-of-computer-science
RA-RAG's 2025 'weighted majority voting' cites neither Condorcet nor Littlestone & Warmuth — but it does inherit WMV from the crowdsourcing literature (Li & Yu 2014)

claim seedling Tier 1 2026-08-10 · RAG, machine-learning-theory, voting-theory, multiple-discovery, source-evaluation
The two axes that won't stay independent — five fields split trust-in-the-source from trust-in-the-message, and the split keeps leaking

topic map 2026-08-10 · source-evaluation, epistemics, intelligence-tradecraft, admiralty-code, RAG
The 2025 intelligence-analysis and RAG findings that evaluators cannot separate source reliability from information credibility are instances of Thorndike's 1920 halo effect, which neither cites

observation seedling Tier 1 2026-08-07 · cross-domain-bridge, cross-time-bridge, cognitive-bias, source-evaluation, intelligence-tradecraft
Thorndike's 1920 halo effect is the century-old mechanism behind the vault's source-reliability/credibility bipartite pair

capture promoted Tier 1 2026-08-03 · epistemics, cognitive-bias, source-evaluation, psychology, history-of-science
Condorcet's 1785 jury theorem is the unstated ancestor of RA-RAG's 'weighted majority voting' — and a 2024 LLM-ensemble study caught the theorem's independence precondition failing exactly where the vault's evaluator-bias note predicts it would

capture promoted Tier 1 2026-08-02 · condorcet, voting-theory, RAG, ensemble-learning, LLM
Lefort et al. (2024) found LLM ensembling produced only marginal sentiment-classification accuracy gains, attributed to non-independent errors that violate the Condorcet Jury Theorem's precondition

claim seedling Tier 1 2026-08-02 · condorcet, voting-theory, LLM, ensemble-learning, epistemics
Evidence-based medicine's GRADE framework independently splits quality of evidence from strength of recommendation, calling their fusion a source of confusion

claim seedling Tier 1 2026-07-21 · vault-design, source-tiers, epistemics, GRADE, cross-domain-bridge
Should the vault's single source_tier field split into two axes — source reliability separate from claim credibility — the way intelligence doctrine and RAG both do?

capture promoted 2026-07-20 · vault-design, source-tiers, provenance, epistemics, meta
The two-axis source grade — from the Admiralty Code to RAG, humans and machines both collapse it

capture promoted 2026-07-11 · intelligence-tradecraft, source-evaluation, provenance, RAG, epistemics
Giving an LLM a webpage's own text degrades its judgment of the source's authority — authority is not textual style or fluency (AuthorityBench)

claim seedling Tier 1 2026-07-11 · RAG, LLM, source-evaluation, authority, epistemics
Reliability-aware RAG estimates each source's reliability separately from document relevance and fuses answers by weighted majority voting

claim seedling Tier 1 2026-07-11 · RAG, source-evaluation, retrieval, LLM, provenance
Cold-War intelligence doctrine and 2025 retrieval research independently converged on the same two-axis source model — reliability apart from credibility — and on the same failure mode

observation seedling Tier 1 2026-07-11 · cross-domain-bridge, cross-time-bridge, source-evaluation, intelligence-tradecraft, RAG
What is AuthorityBench's actual identity — arXiv id, venue, authors — and does the Qwen3-32B ρ 75.28% DomainAuth figure check out?

question closed 2026-07-11 · verification, RAG, source-evaluation, authority, benchmark