index / Maverick-Ansh/intentions_emergent
intentions_emergent
Latent reasoning for user-intent decomposition: does more continuous-space compute produce better goal extraction? With JEV (TypeSafe System One) as the calibrated decision layer.
Research and EvalsUnclearReviewJupyter Notebook
Repository
Stars0
LanguageJupyter Notebook
License—
Created2026-09-18
Last pushed2026-09-18
Officialno
Judgment
Genuine Jev project0.35
gate 0.5
Uses Jev at runtime0.24
Is a meta list0.07
Is a reimplementation0.22
Quality scales
Substance0.24 / 3
gate 0.5
Docs0.00 / 3
Novelty1.67 / 3
Composite score0.20
Category distribution
Research and Evals0.98
Other0.01
Learning0.01
Applications0.00
SDKs and Clients0.00
Agent and Dev Tooling0.00
Integrations0.00
Games and Simulation0.00
Assigned categoryResearch and Evals
Category probability0.98
Category confidence0.97
PatternUnclear
Pattern confidence0.90
Curation
StatusReview
Reasongenuine 0.35 below 0.5
README pickno
Overridenone
Sources
- search:jev typesafe
- search:"system one" jev
Provenance
Modeljev-1.13.0
Question setv2
Judged at2026-09-19 10:54 UTC