{"@context":"https://schema.org","@type":"CreativeWork","@id":"https://forgecascade.org/public/capsules/aa74d3c4-fa93-4fc7-869d-bc1ce1375d4a","identifier":"aa74d3c4-fa93-4fc7-869d-bc1ce1375d4a","url":"https://forgecascade.org/public/capsules/aa74d3c4-fa93-4fc7-869d-bc1ce1375d4a","name":"Quiet-STaR Language Model Reasoning Reference","text":"Zelikman and coauthors present Quiet-STaR, a method for training language models to generate internal rationales before producing tokens, extending the Self-Taught Reasoner line of work. This capsule is a narrow reference for latent reasoning, rationale generation, and inference-time reasoning research. It replaces a private auto-chunk from a broader inference-time compute roundup.\n\nSources:\n- https://arxiv.org/abs/2403.09629","keywords":["arxiv","quiet-star","language-model-reasoning","latent-reasoning","inference-time-compute","manual-public-review","source-backed","public-reference","free-public-reference"],"about":[],"citation":[],"isPartOf":{"@type":"Dataset","name":"Forge Cascade Knowledge Graph","url":"https://forgecascade.org"},"publisher":{"@type":"Organization","name":"Forge Cascade","url":"https://forgecascade.org"},"dateCreated":"2026-05-12T23:43:24.114467Z","dateModified":"2026-06-19T09:56:41.320000Z","isBasedOn":"https://arxiv.org/abs/2403.09629","additionalProperty":[{"@type":"PropertyValue","name":"trust_level","value":90},{"@type":"PropertyValue","name":"verification_status","value":"sources_verified"},{"@type":"PropertyValue","name":"provenance_status","value":"valid"},{"@type":"PropertyValue","name":"evidence_level","value":"primary_source"},{"@type":"PropertyValue","name":"content_hash","value":"51c240c5d2c34c610464dc2999d30a60824e8979aa66e56837f1573cc3f9fce1"}]}