AbstractPhil commited on
Commit
3982b8d
·
verified ·
1 Parent(s): aa2c0f0

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +11 -0
README.md CHANGED
@@ -2,6 +2,17 @@
2
  license: apache-2.0
3
  ---
4
 
 
 
 
 
 
 
 
 
 
 
 
5
  This one is going to attempt a new form of prediction based on velocity flow matching using spherical representation.
6
 
7
  The alignment for the early prototype is essentially going to be euler discreet ODE flow matching, and it will be more than enough for now.
 
2
  license: apache-2.0
3
  ---
4
 
5
+ ```
6
+ Standard: x_t → UNet(x_t, t, c) → v_pred
7
+ ↑ full 16384-dim path
8
+
9
+ GLFM: x_t → encoder → project to S^15 → triangulate against anchors
10
+ ↓
11
+ 768-dim geometric address (continuous, not discrete)
12
+ ↓
13
+ conditioned_generator(address, t, c) → v_pred
14
+ ```
15
+
16
  This one is going to attempt a new form of prediction based on velocity flow matching using spherical representation.
17
 
18
  The alignment for the early prototype is essentially going to be euler discreet ODE flow matching, and it will be more than enough for now.