{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://forgecascade.org/public/capsules/d78feeba-4461-4a9b-a77f-4bb62c1bfa29","identifier":"d78feeba-4461-4a9b-a77f-4bb62c1bfa29","url":"https://forgecascade.org/public/capsules/d78feeba-4461-4a9b-a77f-4bb62c1bfa29","name":"Predictable Confabulations: Factual Recall by LLMs Scales with Model Size and Topic Frequency","text":"# Predictable Confabulations: Factual Recall by LLMs Scales with Model Size and Topic Frequency\n\n**Authors:** Matthew L. Smith, Jonathan P. Shock, Samuel T. Segun, Iyiola E. Olatunji, Tegawendé F. Bissyandé\n**arXiv:** https://arxiv.org/abs/2605.18732v1\n**Published:** 2026-05-18T17:53:44Z\n\n## Abstract\nWhile scaling laws govern aggregate large language model performance, no scaling law has linked factual recall to both model size and training-data composition. We evaluated 38 models on over 8,900 scholarly references evaluated by an automated reference verification system. Recall quality follows a sigmoid in the log-linear combination of model parameter count and topic representation in training data. These two variables alone explain 60% of the variance across 16 dense models from four families, rising to 74-94% within individual families. The form matches a superposition-inspired account in which recall is gated by a signal-to-noise ratio: signal strength scales with concept frequency and the noise floor with model capacity.","keywords":["cs.CL","cs.AI","cs.LG"],"about":[{"@type":"Thing","name":"BlackCat"},{"@type":"Thing","name":"Grandoreiro"},{"@type":"Thing","name":"jRAT"},{"@type":"Thing","name":"Malteiro"},{"@type":"Thing","name":"Play"},{"@type":"Thing","name":"Component Object Model Hijacking"},{"@type":"Thing","name":"Component Object Model"},{"@type":"Thing","name":"Data Transfer Size Limits"}],"citation":[],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://forgecascade.org"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://forgecascade.org"},"dateCreated":"2026-05-19T06:00:07.293000Z","dateModified":"2026-05-19T06:00:07.293000Z","isBasedOn":"https://arxiv.org/abs/2605.18732v1","additionalProperty":[{"@type":"PropertyValue","name":"trust_level","value":65},{"@type":"PropertyValue","name":"verification_status","value":"source_linked"},{"@type":"PropertyValue","name":"evidence_level","value":"primary_source"}]}