Clarify LoRA direction filenames

Rename adapters using apply_to_BASE__toward_TARGET terminology and document compressed-matrix rank fidelity separately from misleading global energy capture.

README.md

@@ -35,7 +35,11 @@ At strength `1.0`, applying an adapter to its matching base moves the model towa
 
 ## Required checkpoints and directions
 
-For files named `ref2va_from_fl2va`, use this base:
+The filenames state both the checkpoint to load and the direction of the
+adapter. `apply_to_fl2va__toward_ref2va` means exactly: load the pruned FL2VA
+checkpoint, then apply that LoRA to move its weights toward Ref2VA.
+
+For files named `apply_to_fl2va__toward_ref2va`, use this base:
 
 ```text
 Comfy-Org/MiniMax-H3
@@ -48,7 +52,7 @@ Base SHA-256:
 a32572fb90b5508b201ec7c2eddcc184b13ddfd3c6f6d2cf06a0b46535d541b4
 ```
 
-For files named `fl2va_from_ref2va`, use this base:
+For files named `apply_to_ref2va__toward_fl2va`, use this base:
 
 ```text
 diffusion_models/minimax_h3_ref2va_pruned_bf16.safetensors
@@ -59,23 +63,41 @@ Do not expect correct results with the full-size BF16, FP8, or INT8 checkpoints.
 
 ## Files
 
-| Direction | File | Size | Weighted matrix-delta energy capture |
-|---|---|---:|---:|
-| FL2VA → Ref2VA | `minimax_h3_ref2va_from_fl2va_pruned_rank256_bf16_lora.safetensors` | 2.589 GB | 99.1260% |
-| FL2VA → Ref2VA | `minimax_h3_ref2va_from_fl2va_pruned_rank512_bf16_lora.safetensors` | 5.075 GB | 99.2417% |
-| FL2VA → Ref2VA | `minimax_h3_ref2va_from_fl2va_pruned_rank1024_bf16_lora.safetensors` | 10.048 GB | 99.4311% |
-| Ref2VA → FL2VA | `minimax_h3_fl2va_from_ref2va_pruned_rank256_bf16_lora.safetensors` | 2.589 GB | 99.1260% |
-| Ref2VA → FL2VA | `minimax_h3_fl2va_from_ref2va_pruned_rank512_bf16_lora.safetensors` | 5.075 GB | 99.2417% |
-| Ref2VA → FL2VA | `minimax_h3_fl2va_from_ref2va_pruned_rank1024_bf16_lora.safetensors` | 10.048 GB | 99.4311% |
-
-The weighted percentage is dominated by high-energy tensors whose effective rank is naturally small. Many large attention and MLP deltas retain substantially less energy, so this statistic is **not** a quality score. Compare generations with identical prompts, inputs, seeds, and sampler settings.
+| Apply to this base | Move toward | File | Size |
+|---|---|---|---:|
+| pruned FL2VA BF16 | Ref2VA | `minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank256.safetensors` | 2.589 GB |
+| pruned FL2VA BF16 | Ref2VA | `minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank512.safetensors` | 5.075 GB |
+| pruned FL2VA BF16 | Ref2VA | `minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank1024.safetensors` | 10.048 GB |
+| pruned Ref2VA BF16 | FL2VA | `minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank256.safetensors` | 2.589 GB |
+| pruned Ref2VA BF16 | FL2VA | `minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank512.safetensors` | 5.075 GB |
+| pruned Ref2VA BF16 | FL2VA | `minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank1024.safetensors` | 10.048 GB |
+
+### What rank preserves
+
+| Rank | Median capture across compressed matrices | Mean capture across compressed matrices | Global energy-weighted capture |
+|---:|---:|---:|---:|
+| 256 | 14.32% | 16.77% | 99.1260% |
+| 512 | 26.40% | 29.70% | 99.2417% |
+| 1024 | 46.11% | 49.62% | 99.4311% |
+
+The global weighted percentage is **not the full story and is not a quality
+score**. It is dominated by a small number of high-energy tensors with low
+effective rank. Of the 264 matrix deltas, 54 low-rank matrices are reconstructed
+essentially exactly at every requested rank. The table's median and mean are
+calculated over the remaining 210 matrices that are genuinely compressed;
+typical large attention and MLP matrices retain much less energy. Rank 1024 can
+therefore be materially more faithful than rank 256 even though their global
+weighted figures differ by only about 0.3 percentage points. Weight-space
+energy also does not predict which directions matter most during inference;
+compare actual generations with identical prompts, inputs, seeds, and sampler
+settings.
 
 Each `.safetensors.json` file contains the complete per-layer extraction report.
 
 ## ComfyUI use
 
 1. Put one `.safetensors` file in `ComfyUI/models/loras/`.
-2. Load pruned FL2VA for a `ref2va_from_fl2va` file, or pruned Ref2VA for a `fl2va_from_ref2va` file.
+2. Follow the filename literally: load pruned FL2VA for an `apply_to_fl2va` file, or pruned Ref2VA for an `apply_to_ref2va` file.
 3. Apply the adapter with a model-only LoRA loader.
 4. Start at model strength `1.0`; compare `0.25`, `0.5`, `0.75`, and `1.0` using identical inputs.
 5. Do not stack ranks or opposing directions together. They are alternative forms of the same underlying delta.

SHA256SUMS

@@ -1,6 +1,6 @@
-f8f15be3728fe52159b2dd840f6c8d7057f99b93690ba5bb129d827757cf13f4  minimax_h3_ref2va_from_fl2va_pruned_rank256_bf16_lora.safetensors
-f7e7fac7e5d29622705f92e1a1cd359db0c54c3364344fdecfc82d8d32b7d512  minimax_h3_ref2va_from_fl2va_pruned_rank512_bf16_lora.safetensors
-b97515fd5e654a3b37c4f499404f73340cb46e93bb626a575cd8626fd889fad0  minimax_h3_ref2va_from_fl2va_pruned_rank1024_bf16_lora.safetensors
-cd4001221599e88f1fa06b2e0cddbc69d61e14765f9b0a994369663f9fc86167  minimax_h3_fl2va_from_ref2va_pruned_rank256_bf16_lora.safetensors
-b7cad943cea4fee3d4614c092f78b370fd3d9cc876f7a5c6f6d254c1aed8b122  minimax_h3_fl2va_from_ref2va_pruned_rank512_bf16_lora.safetensors
-8c90b63df76e1819870c8b6809b08eb2bf8cf80348e59e3117d69e322a481d68  minimax_h3_fl2va_from_ref2va_pruned_rank1024_bf16_lora.safetensors
+f8f15be3728fe52159b2dd840f6c8d7057f99b93690ba5bb129d827757cf13f4  minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank256.safetensors
+f7e7fac7e5d29622705f92e1a1cd359db0c54c3364344fdecfc82d8d32b7d512  minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank512.safetensors
+b97515fd5e654a3b37c4f499404f73340cb46e93bb626a575cd8626fd889fad0  minimax_h3_pruned_bf16__apply_to_fl2va__toward_ref2va__rank1024.safetensors
+cd4001221599e88f1fa06b2e0cddbc69d61e14765f9b0a994369663f9fc86167  minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank256.safetensors
+b7cad943cea4fee3d4614c092f78b370fd3d9cc876f7a5c6f6d254c1aed8b122  minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank512.safetensors
+8c90b63df76e1819870c8b6809b08eb2bf8cf80348e59e3117d69e322a481d68  minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank1024.safetensors

minimax_h3_fl2va_from_ref2va_pruned_rank1024_bf16_lora.safetensors.json → minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank1024.safetensors.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:97009c54e939ff7576cf1c65ede03421eb6035c079d471899543847dc0d127c1
-size 216945
+oid sha256:8ee2f123f3d1a9b25933658437723c8a2c935c6978acd2bcd28c13f46de5bcc5
+size 216963

minimax_h3_fl2va_from_ref2va_pruned_rank256_bf16_lora.safetensors.json → minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank256.safetensors.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f9cc0dccd043471612a5db6f333027e38666ae6b1d57ef07a9ca838c5f3ab6ed
-size 217055
+oid sha256:d342a9d440748acfe9a30a29f16cb3b30f564528984d91c28429683e138a6b5d
+size 217073

minimax_h3_fl2va_from_ref2va_pruned_rank512_bf16_lora.safetensors.json → minimax_h3_pruned_bf16__apply_to_ref2va__toward_fl2va__rank512.safetensors.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05f96c313abfa9e301844dc3ead2b2b7b635c32775d775f366ba27dcf95630ea
-size 216928
+oid sha256:fcaabc6099d36a7f71f5f87a1f4c754af477702a072b070cfb89be6a79ad4481
+size 216946