The plain matrices are not blind to what is stated once
Browse files
README.md
CHANGED
|
@@ -118,7 +118,9 @@ That costs 0.0027 of pure-text cosine on the residual and 0.0013 on the matrix.
|
|
| 118 |
| `mmh3-8b-ClipProj-v3` | 0.9289 |
|
| 119 |
| `mmh3-4b-ClipProj-v3` | 0.9193 |
|
| 120 |
|
| 121 |
-
**Prefer the `-mlp` files.**
|
|
|
|
|
|
|
| 122 |
|
| 123 |
## What this is
|
| 124 |
|
|
|
|
| 118 |
| `mmh3-8b-ClipProj-v3` | 0.9289 |
|
| 119 |
| `mmh3-4b-ClipProj-v3` | 0.9193 |
|
| 120 |
|
| 121 |
+
**Prefer the `-mlp` files.** They measure closer to the 32B — 0.9449 against 0.9289 on the 8B — and on the demo scene both plain matrices get the seated pose wrong: the prompt asks for one bare leg crossed over the other, they sit her cross-legged. It is visible in the first two segments of the video.
|
| 122 |
+
|
| 123 |
+
The gap is narrower than that makes it sound, though, and it is worth being precise about it. The cat, the straw hat, the laundry, the bouncing knee — all of it comes through on the plain matrices too. A linear map is not blind to what is stated once; on this scene it missed one attribute out of many, and it happened to be the pose.
|
| 124 |
|
| 125 |
## What this is
|
| 126 |
|