LLMpedia: A Transparent Framework to Materialize an LLM's Encyclopedic Knowledge at Scale

Abstract

Benchmarks such as MMLU suggest flagship language models approach factuality saturation, with scores above 90\%. We show this picture is incomplete. \emph{LLMpedia} generates encyclopedic articles entirely from parametric memory, producing ${\sim}$1M articles across three model families without retrieval. For gpt-5-mini, the verifiable true rate on Wikipedia-covered subjects is only 74.7\% -- more than 15 percentage points below the benchmark-based picture, consistent with the availability bias of fixed-question evaluation. Beyond Wikipedia, frontier subjects verifiable only through curated web evidence fall further to 63.2\% true rate. Wikipedia covers just 61\% of surfaced subjects, and three model families overlap by only 7.3\% in subject choice. In a capture-trap benchmark inspired by prior analysis of Grokipedia, LLMpedia achieves substantially higher factuality at roughly half the textual similarity to Wikipedia. Unlike Grokipedia, every prompt, artifact, and evaluation verdict is publicly released, making LLMpedia the first fully open parametric encyclopedia -- bridging factuality evaluation and knowledge materialization. All data, code, and a browsable interface are at https://llmpedia.net.

Figures (19)

• Figure 1: LLMpedia generates entirely from parametric memory with full auditability. Grokipedia's opaque pipeline shows evidence of retrieval-shaped generation yasseri2025similar.
• Figure 2: LLMpedia pipeline. Each subject flows through optional self-grounding, dynamic outline generation, and article elicitation. Extracted [[wikilinks]] undergo canonical normalization, LLM-based encyclopedic filtering, and embedding-based deduplication before surviving children enter the BFS queue. Details in Appendix \ref{['app:implementation']}.
• Figure 3: Dynamic outline for Vannevar Bush. Sections adapt to the entity rather than following a fixed template.
• Figure 4: Generated Wikitext excerpt. Blue: retained entities for expansion; red: filtered by Stage 2 (§\ref{['sec:sanitization']}) -- common nouns and generic roles that would produce circular, low-quality sub-articles.
• Figure 5: BFS expansion from seed Vannevar Bush. Blue nodes survive sanitization and enter the queue; red dashed nodes are filtered as generic.