diff options
| author | skal <pascal.massimino@gmail.com> | 2026-02-14 00:50:11 +0100 |
|---|---|---|
| committer | skal <pascal.massimino@gmail.com> | 2026-02-14 00:50:11 +0100 |
| commit | 67ada21b34d87c780f42b1af7958dcf831d864ec (patch) | |
| tree | 44654f744f80fb3b1e78fe7e243ab295f0b37d08 /doc/CNN_V2_DEBUG_TOOLS.md | |
| parent | c7ea4123efa2457e17a9894bba645914c065a190 (diff) | |
gen_identity_weights: Change --mix to 50-50 blend
Updates --mix mode to use 50-50 weighting to avoid overflow:
- Before: p0+p4, p1+p5, p2+p6, p3+p7
- After: 0.5*p0+0.5*p4, 0.5*p1+0.5*p5, etc
Prevents saturation when blending input with static features.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Diffstat (limited to 'doc/CNN_V2_DEBUG_TOOLS.md')
| -rw-r--r-- | doc/CNN_V2_DEBUG_TOOLS.md | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/doc/CNN_V2_DEBUG_TOOLS.md b/doc/CNN_V2_DEBUG_TOOLS.md index dfea63e..8d1289a 100644 --- a/doc/CNN_V2_DEBUG_TOOLS.md +++ b/doc/CNN_V2_DEBUG_TOOLS.md @@ -18,7 +18,7 @@ Tools for investigating CNN v2 mismatch between HTML tool and cnn_test. # 3×3 identity ./training/gen_identity_weights.py workspaces/main/weights/cnn_v2_identity_3x3.bin --kernel-size 3 -# Mix mode: p0+p4, p1+p5, p2+p6, p3+p7 +# Mix mode: 50-50 blend (0.5*p0+0.5*p4, etc) ./training/gen_identity_weights.py output.bin --mix # Static features only: p4→ch0, p5→ch1, p6→ch2, p7→ch3 @@ -31,7 +31,7 @@ Tools for investigating CNN v2 mismatch between HTML tool and cnn_test. **Output:** - Single layer, 12D→4D (4 input channels + 8 static features) - Identity mode: Output Ch{0,1,2,3} = Input Ch{0,1,2,3} -- Mix mode (--mix): Output Ch{i} = Input Ch{i} + Input Ch{i+4} (blends input with static features) +- Mix mode (--mix): Output Ch{i} = 0.5*Input Ch{i} + 0.5*Input Ch{i+4} (50-50 blend, avoids overflow) - Static mode (--p47): Output Ch{i} = Input Ch{i+4} (static features only, visualizes p4-p7) - Minimal file size (~136 bytes for 1×1, ~904 bytes for 3×3) |