Tensor Lab

Explore higher-order tensors produced by the framework. This page is deck-agnostic: it shows the generic overcommit utility tensor and provides exports.

4th-order mode×turn×L×r Asset mask U>0 Download JSON Extensible add more tensors

Overcommit Utility Tensor

Utility: U(mode,L,r) = -Drain(L,r) + ι¡Gain(mode). This models planned tempo debt conversion.

Drain(L,r) = (1-r)L
U = -Drain + ι¡Gain(mode)
Asset iff U > 0

U(mode, turn, L, r)

ModeTurn bucketL r=0.3r=0.4r=0.5r=0.6
STORY early(T3-4) 6 -2.60 -2.00 -1.40 -0.80
7 -3.30 -2.60 -1.90 -1.20
8 -4.00 -3.20 -2.40 -1.60
mid(T5-6) 6 -2.60 -2.00 -1.40 -0.80
7 -3.30 -2.60 -1.90 -1.20
8 -4.00 -3.20 -2.40 -1.60
late(T7+) 6 -2.60 -2.00 -1.40 -0.80
7 -3.30 -2.60 -1.90 -1.20
8 -4.00 -3.20 -2.40 -1.60
HUNT early(T3-4) 6 -2.20 -1.60 -1.00 -0.40
7 -2.90 -2.20 -1.50 -0.80
8 -3.60 -2.80 -2.00 -1.20
mid(T5-6) 6 -2.20 -1.60 -1.00 -0.40
7 -2.90 -2.20 -1.50 -0.80
8 -3.60 -2.80 -2.00 -1.20
late(T7+) 6 -2.20 -1.60 -1.00 -0.40
7 -2.90 -2.20 -1.50 -0.80
8 -3.60 -2.80 -2.00 -1.20
APEX early(T3-4) 6 -1.80 -1.20 -0.60 0.00
7 -2.50 -1.80 -1.10 -0.40
8 -3.20 -2.40 -1.60 -0.80
mid(T5-6) 6 -1.80 -1.20 -0.60 0.00
7 -2.50 -1.80 -1.10 -0.40
8 -3.20 -2.40 -1.60 -0.80
late(T7+) 6 -1.80 -1.20 -0.60 0.00
7 -2.50 -1.80 -1.10 -0.40
8 -3.20 -2.40 -1.60 -0.80
Download tensor JSON α 2 Gain STORY 0.8 • HUNT 1 • APEX 1.2

Add your own tensors

The framework supports additional tensors without changing the UI contract: export them into the run JSON and render via tables or charts.

Examples:
- color×turn×pip requirements
- removal density×matchup class
- land source tensor (special lands Phase 2)
- narrative coherence tensor (lore tags)

Design principle

Tensors are how you make “repeatable tuning decisions”: you stop arguing with your mood and start arguing with a matrix.