SOURCE-LINKED INTELLIGENCE

Specification Oracles

arXiv · AI, language, vision and robotics · article · Sep 11, 2026 · UTC

Specifications face a basic tradeoff: leave details out, and important questions go unanswered; record every detail separately, and the specification becomes large and prolix. We investigate whether a language model can serve as a compact, living specification oracle by learning facts about a target and answering questions about it directly. We compare two ways of storing the learned facts: external text notes and changes to the model's weights. Across four families of 596-fact worlds and two Qwen2.5 model sizes, weight-only oracles benefited substantially more from structure: with the 7B mode

Read original source ↗ Open in workspace

recordType

paper

region

Global