Awesome Jev

index / j1s4nn/JevGAG

JevGAG

JevGAG + DDAG: a two-GPU study of adaptive temporal granularity for causal video generation and video world models - the JevGAG report (preconditions) and the DDAG method (regime map, intrinsic probe, seam repair), with a boundary analysis.

Research and EvalsRoutingReviewPython

Repository

Stars1
LanguagePython
LicenseMIT
Created2026-10-25
Last pushed2026-10-28
Officialno

Judgment

Genuine Jev project0.41
gate 0.5
Uses Jev at runtime0.24
Is a meta list0.04
Is a reimplementation0.25

Quality scales

Substance2.05 / 3
gate 0.5
Docs2.49 / 3
Novelty2.61 / 3
Composite score0.79

Category distribution

Research and Evals0.86
Other0.13
Games and Simulation0.01
SDKs and Clients0.00
Integrations0.00
Agent and Dev Tooling0.00
Applications0.00
Learning0.00
Assigned categoryResearch and Evals
Category probability0.86
Category confidence0.84
PatternRouting
Pattern confidence0.18

Curation

StatusReview
Reasongenuine 0.41 below 0.5
README pickno
Overridenone

Sources

Provenance

Modeljev-1.14.0
Question setv2
Judged at2026-11-01 12:33 UTC