Force-Intent Model — Cortix AI

[ 001 ] Force-Intent Model

What the body feels, before it moves.

A single latent model that fuses what the body feels, what the operator intends, and what the world will do next.

Sensor inputs · x_t

EOGGaze

EMGMuscle

IMUMotion

TACForce

VISVision

fuse

→

Latent world model · z_t

Force × Intent × World

z_t = f(EOG, EMG, IMU, TAC, VIS)
z_t+1 = T(z_t, a_t)

Perceive

Predict

Act

decode

→

Model outputs · y_t

F_tPredicted force

I_tInferred intent

S_t+1Next state

π_tAction policy

Force

What the body exerts and receives — measured, not guessed.

Intent

Where attention, muscle, and gaze converge before motion begins.

World

A rollable simulator of the next moment — physically consistent.