diff --git a/README.md b/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..510b81ce6363b2bb929e64ee44c004b7794022ba
--- /dev/null
+++ b/README.md
@@ -0,0 +1,91 @@
+---
+license: apache-2.0
+tags:
+- image-classification
+- probing
+- frozen-features
+- imagenet-1k
+library_name: pytorch
+---
+
+# Efficient Probing heads
+
+Trained EP (efficient probing) heads for 36 frozen encoders, from the standing
+ImageNet-1k benchmark at [billpsomas/efficient-probing](https://github.com/billpsomas/efficient-probing)
+(ICLR 2026). Each head is a multi-query cross-attention pooling + BatchNorm + linear
+classifier, trained for 90 epochs with LARS on frozen features. **No backbone weights
+are included** -- each `config.json` records the exact `--model` / loader flags to
+rebuild the encoder from its original source.
+
+Every head ships with its accuracy **at the saved epoch** in `meta`/`config.json`:
+12 heads are the run's peak (they reproduce the leaderboard number exactly); the rest
+are final-epoch heads, with both figures recorded so nothing is ambiguous. DiT-XL/2
+is absent (no surviving checkpoint; see the manifest).
+
+## Loading
+
+```python
+from huggingface_hub import hf_hub_download
+import torch
+
+path = hf_hub_download("billpsomas/efficient-probing-heads",
+                       "dinov3_vit7b/ep_head.pth")
+ck = torch.load(path, map_location="cpu", weights_only=False)
+head_state, meta = ck["state_dict"], ck["meta"]
+# with the benchmark repo on PYTHONPATH:
+#   model = backbones.build_backbone(args, device)   # args from meta
+#   probe_heads.build_probe_head(model, args)
+#   model.head.load_state_dict(head_state, strict=True)
+```
+
+Or evaluate directly with the benchmark's tool:
+
+```
+python tools/eval_reimagenet.py predict <backbone flags from config.json> \
+    --head_ckpt ep_head.pth --pred_out preds.json
+```
+
+## Heads
+
+| encoder | EP variant | top-1 @ saved epoch | epoch | checkpoint is |
+|---|---|---:|---:|---|
+| DINOv3 ViT-7B/16 | ep_all | 88.36 | 6 | peak |
+| MetaCLIP2 ViT-bigG/14-378 | ep | 88.12 | 6 | peak |
+| EVA02-CLIP E-14-plus | ep | 87.98 | 6 | peak |
+| EVA02-CLIP E-14 | ep | 87.70 | 6 | peak |
+| SigLIP2 SO400M/14 | ep | 85.64 | 29 | final epoch |
+| PE-Core L-14/336 | ep | 87.25 | 12 | peak |
+| MetaCLIP2 ViT-bigG/14 | ep | 87.11 | 6 | peak |
+| SigLIP2 ViT-L/16 | ep | 87.06 | 6 | peak |
+| DINOv3 ViT-L/16 | ep_all | 86.73 | 19 | final epoch |
+| AIMv2 ViT-L/14 | ep | 85.62 | 19 | final epoch |
+| SigLIP ViT-L/16 | ep | 85.93 | 6 | peak |
+| DINOv2 ViT-L/14 | ep_all | 85.56 | 15 | peak |
+| Franca ViT-L/14 | ep_all | 84.28 | 14 | peak |
+| DINOv3 ViT-B/16 | ep_all | 83.77 | 20 | final epoch |
+| DINOv2 ViT-B/14 | ep | 83.61 | 25 | final epoch |
+| RADIO ViT-L/16 | ep | 83.40 | 89 | final epoch |
+| EVA02 ViT-L/14 | ep | 83.22 | 89 | final epoch |
+| CLIP ViT-L/14 | ep | 83.22 | 11 | peak |
+| CAPI ViT-L/14 | ep | 82.43 | 89 | final epoch |
+| BEiTv2 ViT-B/16 | ep | 81.32 | 89 | final epoch |
+| RADIO ViT-B/16 | ep | 80.26 | 89 | final epoch |
+| iBOT ViT-L/16 | ep_all | 79.43 | 89 | final epoch |
+| Hiera ViT-H/16 | ep | 79.82 | 89 | final epoch |
+| MAE ViT-L/16 | ep | 79.43 | 89 | final epoch |
+| I-JEPA ViT-H/14 | ep | 78.80 | 89 | final epoch |
+| iBOT ViT-B/16 | ep_all | 78.62 | 89 | final epoch |
+| Hiera ViT-L/16 | ep | 78.51 | 83 | final epoch |
+| CLIP ViT-B/16 | ep_all | 77.85 | 11 | peak |
+| DINO ViT-B/16 | ep_all | 77.08 | 89 | final epoch |
+| MoCov3 ViT-B/16 | ep_all | 76.21 | 89 | final epoch |
+| Hiera ViT-B/16 | ep | 75.63 | 88 | final epoch |
+| MAE ViT-B/16 | ep | 75.35 | 86 | final epoch |
+| MaskFeat ViT-B/16 | ep | 71.68 | 89 | final epoch |
+| MaskFeat ViT-L/16 | ep | 69.56 | 89 | final epoch |
+| SimMIM ViT-B/16 | ep | 64.81 | 89 | final epoch |
+| MAE ViT-S/16 | ep | 64.56 | 89 | final epoch |
+
+Full provenance (training logs, exact commands, the leaderboard itself) lives in the
+[GitHub repo](https://github.com/billpsomas/efficient-probing). Heads were trained on
+ImageNet-1k; use accordingly.
diff --git a/aimv2_vitl/config.json b/aimv2_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..46cfafe625045aa8adfc0d783e7499f86d157b44
--- /dev/null
+++ b/aimv2_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "AIMv2",
+ "arch": "ViT-L/14",
+ "pretrain": "DFN-2B + COYO + HQITP",
+ "image_size": 224,
+ "model": "aimv2-large-patch14-224",
+ "loader": "aimv2",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 19,
+ "val_acc1_at_head_epoch": 85.62,
+ "table_best_acc1": 85.94,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/aimv2_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/aimv2_vitl/ep_head.pth b/aimv2_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d45ac923b82aa91c4f2d0cd354ec511f9d2033eb
--- /dev/null
+++ b/aimv2_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7565e9ccc7791a3cdcb2ae6cffd2fa2650f2d994807a6b5415de875e4d9bc41
+size 8437201
diff --git a/beitv2_vitb/config.json b/beitv2_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..816d75f893e95b65074961e9abd023b8c3ea94cf
--- /dev/null
+++ b/beitv2_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "BEiTv2",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "simmim",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.beitv2/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 81.32,
+ "table_best_acc1": 81.39,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/beitv2_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/beitv2_vitb/ep_head.pth b/beitv2_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9ac5d6648e942395769521a61f25f7de13c73449
--- /dev/null
+++ b/beitv2_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98ff601343eb6f603dac0da5171a19871890274c1d3e9b76b28fae7951445250
+size 5543441
diff --git a/capi_vitl/config.json b/capi_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6a5c27a5a0021ba30c138689e25f787ae4ce32f
--- /dev/null
+++ b/capi_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "CAPI",
+ "arch": "ViT-L/14",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "capi_vitl14_in1k",
+ "loader": "capi",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 82.43,
+ "table_best_acc1": 82.98,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/capi_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/capi_vitl/ep_head.pth b/capi_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..cb871add40a559fa416fe72bc3f6b59f1372e6e9
--- /dev/null
+++ b/capi_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce826a1e518624a416e0a8887c4532170a271d069bf1176287d9bca633bc58f6
+size 8437201
diff --git a/clip_vitb16/config.json b/clip_vitb16/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..18c897d1372ad1f0afbd28a496ad4bb418bbab8b
--- /dev/null
+++ b/clip_vitb16/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "CLIP",
+ "arch": "ViT-B/16",
+ "pretrain": "WIT-400M",
+ "image_size": 224,
+ "model": "ViT-B-16",
+ "loader": "openclip",
+ "pretrain_tag": "openai",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 11,
+ "val_acc1_at_head_epoch": 77.85,
+ "table_best_acc1": 77.85,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/clip_vitb16/ep.txt"
+}
\ No newline at end of file
diff --git a/clip_vitb16/ep_head.pth b/clip_vitb16/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..41925ac7a2ac0141a112a3ce1773edea5174ffcc
--- /dev/null
+++ b/clip_vitb16/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ccc806165e35b6b2bf24e5d969168700f1445a8fd9b4a75b54ecf106bc87310
+size 3173713
diff --git a/clip_vitl/config.json b/clip_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..bd103b2502564bd3fa3bf1a4a834ca28deebcca7
--- /dev/null
+++ b/clip_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "CLIP",
+ "arch": "ViT-L/14",
+ "pretrain": "WIT-400M",
+ "image_size": 224,
+ "model": "ViT-L-14",
+ "loader": "openclip",
+ "pretrain_tag": "openai",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 11,
+ "val_acc1_at_head_epoch": 83.22,
+ "table_best_acc1": 83.22,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/clip_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/clip_vitl/ep_head.pth b/clip_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0e5f73cb01f39a91eca15f12ebfb1fc464b51b8d
--- /dev/null
+++ b/clip_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bd6460e5b20f6b7355a8522e7034d195a6000127dd6b9af8591cde2ef0b7515
+size 8437073
diff --git a/dino_vitb/config.json b/dino_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d287692a44d5a0b759fb3bf9265736ec9bf17c46
--- /dev/null
+++ b/dino_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINO",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.dino",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 77.08,
+ "table_best_acc1": 77.39,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/dino_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/dino_vitb/ep_head.pth b/dino_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a7bf4420e04644a94a4c0a4aef7bd2c23c168911
--- /dev/null
+++ b/dino_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76941d4fb977f8b2c21b998c805ef37935886c778b0ad12b6a7c1edd904cce10
+size 5543377
diff --git a/dinov2_vitb/config.json b/dinov2_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5f6e3b2f37ea7d72f49c12ce6db311394168bf9
--- /dev/null
+++ b/dinov2_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINOv2",
+ "arch": "ViT-B/14",
+ "pretrain": "LVD-142M",
+ "image_size": 224,
+ "model": "dinov2_vitb14",
+ "loader": "hub",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 25,
+ "val_acc1_at_head_epoch": 83.61,
+ "table_best_acc1": 84.01,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/dinov2_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/dinov2_vitb/ep_head.pth b/dinov2_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d2a9b410a8e6ec2f822cf5d06a67728a68222997
--- /dev/null
+++ b/dinov2_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13bcf74040d8d47cfca8f7eea7ef80b8391ab14bbd0fe1a1f94b4ad7b1687668
+size 5543377
diff --git a/dinov2_vitl/config.json b/dinov2_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..66431a1e51aa9169c0f8be0f9b819473a2d5df57
--- /dev/null
+++ b/dinov2_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINOv2",
+ "arch": "ViT-L/14",
+ "pretrain": "LVD-142M",
+ "image_size": 224,
+ "model": "dinov2_vitl14",
+ "loader": "hub",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 15,
+ "val_acc1_at_head_epoch": 85.56,
+ "table_best_acc1": 85.56,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/dinov2_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/dinov2_vitl/ep_head.pth b/dinov2_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..5c9d73a1327f2942dc2162c0c9e3220cb2c76331
--- /dev/null
+++ b/dinov2_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d72ad24de49a8af0eddfc062203a32f4536a5a2871ccf9fad5c6f0c610e868e
+size 8437073
diff --git a/dinov3_vit7b/config.json b/dinov3_vit7b/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..af8ee86ab01f02865c0a105f9122ea3af9cac85a
--- /dev/null
+++ b/dinov3_vit7b/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINOv3",
+ "arch": "ViT-7B/16",
+ "pretrain": "LVD-1689M",
+ "image_size": 224,
+ "model": "dinov3_vit7b16",
+ "loader": "dinov3",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 88.36,
+ "table_best_acc1": 88.36,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/dinov3_vit7b/ep.txt"
+}
\ No newline at end of file
diff --git a/dinov3_vit7b/ep_head.pth b/dinov3_vit7b/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..dff9f3d96ef00238dc3938823e86b47ac70cef2a
--- /dev/null
+++ b/dinov3_vit7b/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7687980bc0cba8f25e9157bfb8279efc4f67487ef787c3a36c5995a133add666
+size 84057425
diff --git a/dinov3_vitb/config.json b/dinov3_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..9db8b49c014361fd9a65b3ed4f6ddb981721d922
--- /dev/null
+++ b/dinov3_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINOv3",
+ "arch": "ViT-B/16",
+ "pretrain": "LVD-1689M",
+ "image_size": 224,
+ "model": "dinov3_vitb16",
+ "loader": "dinov3",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 20,
+ "val_acc1_at_head_epoch": 83.77,
+ "table_best_acc1": 84.11,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/dinov3_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/dinov3_vitb/ep_head.pth b/dinov3_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..34bf81f1725d0191477fc98b59359d803865b034
--- /dev/null
+++ b/dinov3_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a55d603fefc8eec3bc1fb6ecca75126ee6593303b5d382e2dd72e8a6fc57245a
+size 5543377
diff --git a/dinov3_vitl/config.json b/dinov3_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d8f05bb63b1ea693e31e4f4c15f57f9c153c6a2f
--- /dev/null
+++ b/dinov3_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "DINOv3",
+ "arch": "ViT-L/16",
+ "pretrain": "LVD-1689M",
+ "image_size": 224,
+ "model": "dinov3_vitl16",
+ "loader": "dinov3",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 19,
+ "val_acc1_at_head_epoch": 86.73,
+ "table_best_acc1": 87.05,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/dinov3_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/dinov3_vitl/ep_head.pth b/dinov3_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..2b35c398631e83ef001f3ec5f681e952a9df903d
--- /dev/null
+++ b/dinov3_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:718a1ac2802741c31199265fa8e501a40cf91eaf156e269837d8a0ba5ffc6040
+size 8437201
diff --git a/eva02_e14/config.json b/eva02_e14/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..03a16932eeb63046a3c4f3d5ac6f1addd429279a
--- /dev/null
+++ b/eva02_e14/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "EVA02-CLIP",
+ "arch": "E-14",
+ "pretrain": "LAION-2B (4B seen)",
+ "image_size": 224,
+ "model": "EVA02-E-14",
+ "loader": "openclip",
+ "pretrain_tag": "laion2b_s4b_b115k",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 87.7,
+ "table_best_acc1": 87.7,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/eva02_e14/ep.txt"
+}
\ No newline at end of file
diff --git a/eva02_e14/ep_head.pth b/eva02_e14/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..759bf125f0741c229ebe2605a3ee5396350507d9
--- /dev/null
+++ b/eva02_e14/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5bc9dd3fa66bc9c7a5a38287d02adf004765544aa935dc3b16aeab1da1456b68
+size 20264273
diff --git a/eva02_e14_plus/config.json b/eva02_e14_plus/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7c126f63b0b51157db7d1a620fdcf87ca8f1f3e
--- /dev/null
+++ b/eva02_e14_plus/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "EVA02-CLIP",
+ "arch": "E-14-plus",
+ "pretrain": "LAION-2B (9B seen)",
+ "image_size": 224,
+ "model": "EVA02-E-14-plus",
+ "loader": "openclip",
+ "pretrain_tag": "laion2b_s9b_b144k",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 87.98,
+ "table_best_acc1": 87.98,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/eva02_e14_plus/ep.txt"
+}
\ No newline at end of file
diff --git a/eva02_e14_plus/ep_head.pth b/eva02_e14_plus/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9a3a4f6520d1522e2c883d780eca4d70dfabe0d9
--- /dev/null
+++ b/eva02_e14_plus/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:868d4b18ed28acfab9871585d9eeb7ce61d56b60589e378f9dddc0b018127062
+size 20264337
diff --git a/eva02_vitl_mim/config.json b/eva02_vitl_mim/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe0d459071874a3055e73a1083e2bbc4af36940c
--- /dev/null
+++ b/eva02_vitl_mim/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "EVA02",
+ "arch": "ViT-L/14",
+ "pretrain": "IN-22K",
+ "image_size": 224,
+ "model": "eva02_large_patch14_224.mim_in22k",
+ "loader": "timm",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 83.22,
+ "table_best_acc1": 83.58,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/eva02_vitl_mim/ep.txt"
+}
\ No newline at end of file
diff --git a/eva02_vitl_mim/ep_head.pth b/eva02_vitl_mim/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c81d4cd89c7270c0d8f589754e9ff20a663ac1be
--- /dev/null
+++ b/eva02_vitl_mim/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b945f2cf9c040e65cac57bffe8a79b211e6dec6b889f507423b591897cd09d
+size 8437201
diff --git a/franca_laion/config.json b/franca_laion/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a018443a0463e9f15fe32c7592e87d420e806aff
--- /dev/null
+++ b/franca_laion/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "Franca",
+ "arch": "ViT-L/14",
+ "pretrain": "LAION-600M",
+ "image_size": 224,
+ "model": "franca_vitl14",
+ "loader": "franca",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "--franca_weights LAION",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 14,
+ "val_acc1_at_head_epoch": 84.28,
+ "table_best_acc1": 84.28,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/franca_laion/ep_all.txt"
+}
\ No newline at end of file
diff --git a/franca_laion/ep_head.pth b/franca_laion/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..33f5d464dce0daac6bb95e4ef46249da46378f0b
--- /dev/null
+++ b/franca_laion/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f930b15a1f4a11f723c943b03e2bc83283730e4619c792fcec6e3c07af6eea6
+size 8437073
diff --git a/hiera_base/config.json b/hiera_base/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f52d7ee5df60c315c12d3cb72dd8ae1d631e1095
--- /dev/null
+++ b/hiera_base/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "Hiera",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "hiera_base_224.mae",
+ "loader": "timm",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 88,
+ "val_acc1_at_head_epoch": 75.63,
+ "table_best_acc1": 75.68,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/hiera_base/ep.txt"
+}
\ No newline at end of file
diff --git a/hiera_base/ep_head.pth b/hiera_base/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..03240ea60e9ab04a5b025d51938941c9e8caeea4
--- /dev/null
+++ b/hiera_base/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c023a84619b4bd5b22f043e9ca4d14afbcb608975f23eb7d9c6a676317304d2d
+size 5543377
diff --git a/hiera_huge/config.json b/hiera_huge/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3d80f457077a351bf487f87a37b15194a1770c6
--- /dev/null
+++ b/hiera_huge/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "Hiera",
+ "arch": "ViT-H/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "hiera_huge_224.mae",
+ "loader": "timm",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 79.82,
+ "table_best_acc1": 79.89,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/hiera_huge/ep.txt"
+}
\ No newline at end of file
diff --git a/hiera_huge/ep_head.pth b/hiera_huge/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0ba8c5baf7745e4ad0a2cfda0e2c5fe51cddca4a
--- /dev/null
+++ b/hiera_huge/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:183f0f8dcb99bc13576410a06968c6a9040a1ce2ce2c6b931141bcd01dbc9836
+size 25255377
diff --git a/hiera_large/config.json b/hiera_large/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa8474ffff81c28d7155e785f44895100b93bd51
--- /dev/null
+++ b/hiera_large/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "Hiera",
+ "arch": "ViT-L/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "hiera_large_224.mae",
+ "loader": "timm",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 83,
+ "val_acc1_at_head_epoch": 78.51,
+ "table_best_acc1": 78.55,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/hiera_large/ep.txt"
+}
\ No newline at end of file
diff --git a/hiera_large/ep_head.pth b/hiera_large/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1c96d95093588aea6a839582df4cdddbfd54fd3a
--- /dev/null
+++ b/hiera_large/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffd14707b8652a2725a6788cb57b8b2c3ef04d7c29f501a49048ae13eb0e8f03
+size 10080721
diff --git a/ibot_vitb/config.json b/ibot_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..df2b2c31e2c4e5d6462eb91354f74a10dc5f4fd6
--- /dev/null
+++ b/ibot_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "iBOT",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.ibot/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 78.62,
+ "table_best_acc1": 78.98,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/ibot_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/ibot_vitb/ep_head.pth b/ibot_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..4135f42e62fc44ba65b7576ed70585cc21340f40
--- /dev/null
+++ b/ibot_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b21c07f716b4aa0c0274b64478e2e98ebb412597206627321cca81ae685c363d
+size 5543441
diff --git a/ibot_vitl/config.json b/ibot_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5d7e8154365223a6e16765e92b83bbc902e4df43
--- /dev/null
+++ b/ibot_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "iBOT",
+ "arch": "ViT-L/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_large_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_large_patch16_224.ibot/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 79.43,
+ "table_best_acc1": 79.97,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/ibot_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/ibot_vitl/ep_head.pth b/ibot_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b7aad2b03dceec0ea283082f39a62bff444000ec
--- /dev/null
+++ b/ibot_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:839dcfeabe8eff1868d73e6d4111d6ba6c6496f19e66cc3c8bc2949110c2206f
+size 8437265
diff --git a/ijepa_vith/config.json b/ijepa_vith/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..9632af2a18ed9895e26d474779c607d377c54691
--- /dev/null
+++ b/ijepa_vith/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "I-JEPA",
+ "arch": "ViT-H/14",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_huge_patch14",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_huge_patch14_224.ijepa/checkpoint-799.pth",
+ "extra_flags": "--no_cls_token",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 78.8,
+ "table_best_acc1": 79.0,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/ijepa_vith/ep.txt"
+}
\ No newline at end of file
diff --git a/ijepa_vith/ep_head.pth b/ijepa_vith/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..d02d010ab86210e7acc31ba367874e578ba5a275
--- /dev/null
+++ b/ijepa_vith/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1d74dcc70511357a6911785ddb7983a3f19368a1a497b62c618b80260cad966
+size 11855377
diff --git a/mae_base/config.json b/mae_base/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c2b4a9fcc481454c7a9760daad3fcc9e28c56e3a
--- /dev/null
+++ b/mae_base/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MAE",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.mae",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 86,
+ "val_acc1_at_head_epoch": 75.35,
+ "table_best_acc1": 75.46,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/mae_base/ep.txt"
+}
\ No newline at end of file
diff --git a/mae_base/ep_head.pth b/mae_base/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..58c9edce07bc8dc00352603dba4125dd3a700155
--- /dev/null
+++ b/mae_base/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9425963f1a1575305694ca1875bed9f80140e53cd263548cd81c6d68fbacb1c7
+size 5543377
diff --git a/mae_vitl/config.json b/mae_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..21e444ae80346635f5ecd43a221fff3e60a1430e
--- /dev/null
+++ b/mae_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MAE",
+ "arch": "ViT-L/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_large_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_large_patch16_224.mae",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 79.43,
+ "table_best_acc1": 79.52,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/mae_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/mae_vitl/ep_head.pth b/mae_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..62340c08f6201d8e9ee1efa924a99b3d44fe7b70
--- /dev/null
+++ b/mae_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3935dbcf3cf125e6d2c4f6de070bac04c250c0fa15bb493c6d12d1e3ea6e151
+size 8437201
diff --git a/mae_vits/config.json b/mae_vits/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..2044554e5f26f61cbfccc5b3ae3f7576d291150f
--- /dev/null
+++ b/mae_vits/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MAE",
+ "arch": "ViT-S/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_small_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_small_patch16_224.mae/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 64.56,
+ "table_best_acc1": 64.58,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/mae_vits/ep.txt"
+}
\ No newline at end of file
diff --git a/mae_vits/ep_head.pth b/mae_vits/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..f21dba2112a00444306107a6d142d12185349d62
--- /dev/null
+++ b/mae_vits/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ba33f00307a72367c0a560dc36b364a24f0a15f2f55b8c1a7120c1d0dcadf1b
+size 2185681
diff --git a/manifest.json b/manifest.json
new file mode 100644
index 0000000000000000000000000000000000000000..9dfd524fe753071d4ffe38428a037d0d4b48308b
--- /dev/null
+++ b/manifest.json
@@ -0,0 +1,803 @@
+{
+ "heads": [
+  {
+   "method": "MAE",
+   "arch": "ViT-S/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_small_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_small_patch16_224.mae/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 64.56,
+   "table_best_acc1": 64.58,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/mae_vits/ep.txt",
+   "params_incl_bn_stats": 545513,
+   "file": "mae_vits/ep_head.pth",
+   "size_mb": 2.2
+  },
+  {
+   "method": "MAE",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.mae",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 86,
+   "val_acc1_at_head_epoch": 75.35,
+   "table_best_acc1": 75.46,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/mae_base/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "mae_base/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "MAE",
+   "arch": "ViT-L/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_large_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_large_patch16_224.mae",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 79.43,
+   "table_best_acc1": 79.52,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/mae_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "mae_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "BEiTv2",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "simmim",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.beitv2/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 81.32,
+   "table_best_acc1": 81.39,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/beitv2_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "beitv2_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "SimMIM",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "simmim",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.simmim/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 64.81,
+   "table_best_acc1": 64.86,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/simmim_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "simmim_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "CAPI",
+   "arch": "ViT-L/14",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "capi_vitl14_in1k",
+   "loader": "capi",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 82.43,
+   "table_best_acc1": 82.98,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/capi_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "capi_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "DINO",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.dino",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 77.08,
+   "table_best_acc1": 77.39,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/dino_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "dino_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "iBOT",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.ibot/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 78.62,
+   "table_best_acc1": 78.98,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/ibot_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "ibot_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "DINOv2",
+   "arch": "ViT-B/14",
+   "pretrain": "LVD-142M",
+   "image_size": 224,
+   "model": "dinov2_vitb14",
+   "loader": "hub",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 25,
+   "val_acc1_at_head_epoch": 83.61,
+   "table_best_acc1": 84.01,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/dinov2_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "dinov2_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "DINOv2",
+   "arch": "ViT-L/14",
+   "pretrain": "LVD-142M",
+   "image_size": 224,
+   "model": "dinov2_vitl14",
+   "loader": "hub",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 15,
+   "val_acc1_at_head_epoch": 85.56,
+   "table_best_acc1": 85.56,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/dinov2_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "dinov2_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "DINOv3",
+   "arch": "ViT-B/16",
+   "pretrain": "LVD-1689M",
+   "image_size": 224,
+   "model": "dinov3_vitb16",
+   "loader": "dinov3",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 20,
+   "val_acc1_at_head_epoch": 83.77,
+   "table_best_acc1": 84.11,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/dinov3_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "dinov3_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "DINOv3",
+   "arch": "ViT-L/16",
+   "pretrain": "LVD-1689M",
+   "image_size": 224,
+   "model": "dinov3_vitl16",
+   "loader": "dinov3",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 19,
+   "val_acc1_at_head_epoch": 86.73,
+   "table_best_acc1": 87.05,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/dinov3_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "dinov3_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "CLIP",
+   "arch": "ViT-L/14",
+   "pretrain": "WIT-400M",
+   "image_size": 224,
+   "model": "ViT-L-14",
+   "loader": "openclip",
+   "pretrain_tag": "openai",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 11,
+   "val_acc1_at_head_epoch": 83.22,
+   "table_best_acc1": 83.22,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/clip_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "clip_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "SigLIP",
+   "arch": "ViT-L/16",
+   "pretrain": "WebLI (10B imgs, 109 langs)",
+   "image_size": 256,
+   "model": "ViT-L-16-SigLIP-256",
+   "loader": "openclip",
+   "pretrain_tag": "webli",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 85.93,
+   "table_best_acc1": 85.93,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/siglip_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "siglip_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "SigLIP2",
+   "arch": "ViT-L/16",
+   "pretrain": "WebLI (10B imgs, 109 langs)",
+   "image_size": 256,
+   "model": "ViT-L-16-SigLIP2-256",
+   "loader": "openclip",
+   "pretrain_tag": "webli",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 87.06,
+   "table_best_acc1": 87.06,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/siglip2_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "siglip2_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "AIMv2",
+   "arch": "ViT-L/14",
+   "pretrain": "DFN-2B + COYO + HQITP",
+   "image_size": 224,
+   "model": "aimv2-large-patch14-224",
+   "loader": "aimv2",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 19,
+   "val_acc1_at_head_epoch": 85.62,
+   "table_best_acc1": 85.94,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/aimv2_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "aimv2_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "SigLIP2",
+   "arch": "SO400M/14",
+   "pretrain": "WebLI (10B imgs, 109 langs)",
+   "image_size": 224,
+   "model": "ViT-SO400M-14-SigLIP2",
+   "loader": "openclip",
+   "pretrain_tag": "webli",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 29,
+   "val_acc1_at_head_epoch": 85.64,
+   "table_best_acc1": 87.68,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/siglip2_so400m/ep.txt",
+   "params_incl_bn_stats": 2519273,
+   "file": "siglip2_so400m/ep_head.pth",
+   "size_mb": 10.1
+  },
+  {
+   "method": "EVA02",
+   "arch": "ViT-L/14",
+   "pretrain": "IN-22K",
+   "image_size": 224,
+   "model": "eva02_large_patch14_224.mim_in22k",
+   "loader": "timm",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 83.22,
+   "table_best_acc1": 83.58,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/eva02_vitl_mim/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "eva02_vitl_mim/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "RADIO",
+   "arch": "ViT-L/16",
+   "pretrain": "distilled (DFN5B/SigLIP/DINOv2/SAM)",
+   "image_size": 224,
+   "model": "radio_v2.5-l",
+   "loader": "radio",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 83.4,
+   "table_best_acc1": 83.89,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/radio_v25_l/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "radio_v25_l/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "Hiera",
+   "arch": "ViT-H/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "hiera_huge_224.mae",
+   "loader": "timm",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 79.82,
+   "table_best_acc1": 79.89,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/hiera_huge/ep.txt",
+   "params_incl_bn_stats": 6312937,
+   "file": "hiera_huge/ep_head.pth",
+   "size_mb": 25.3
+  },
+  {
+   "method": "Hiera",
+   "arch": "ViT-L/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "hiera_large_224.mae",
+   "loader": "timm",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 83,
+   "val_acc1_at_head_epoch": 78.51,
+   "table_best_acc1": 78.55,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/hiera_large/ep.txt",
+   "params_incl_bn_stats": 2519273,
+   "file": "hiera_large/ep_head.pth",
+   "size_mb": 10.1
+  },
+  {
+   "method": "Hiera",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "hiera_base_224.mae",
+   "loader": "timm",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 88,
+   "val_acc1_at_head_epoch": 75.63,
+   "table_best_acc1": 75.68,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/hiera_base/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "hiera_base/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "RADIO",
+   "arch": "ViT-B/16",
+   "pretrain": "distilled (DFN5B/SigLIP/DINOv2/SAM)",
+   "image_size": 224,
+   "model": "radio_v2.5-b",
+   "loader": "radio",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 80.26,
+   "table_best_acc1": 80.29,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/radio_v25_b/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "radio_v25_b/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "CLIP",
+   "arch": "ViT-B/16",
+   "pretrain": "WIT-400M",
+   "image_size": 224,
+   "model": "ViT-B-16",
+   "loader": "openclip",
+   "pretrain_tag": "openai",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 11,
+   "val_acc1_at_head_epoch": 77.85,
+   "table_best_acc1": 77.85,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/clip_vitb16/ep.txt",
+   "params_incl_bn_stats": 792553,
+   "file": "clip_vitb16/ep_head.pth",
+   "size_mb": 3.2
+  },
+  {
+   "method": "PE-Core",
+   "arch": "L-14/336",
+   "pretrain": "MetaCLIP-curated 5.4B",
+   "image_size": 336,
+   "model": "PE-Core-L-14-336",
+   "loader": "openclip",
+   "pretrain_tag": "meta",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 12,
+   "val_acc1_at_head_epoch": 87.25,
+   "table_best_acc1": 87.25,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/pe_core_l336/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "pe_core_l336/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "EVA02-CLIP",
+   "arch": "E-14",
+   "pretrain": "LAION-2B (4B seen)",
+   "image_size": 224,
+   "model": "EVA02-E-14",
+   "loader": "openclip",
+   "pretrain_tag": "laion2b_s4b_b115k",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 87.7,
+   "table_best_acc1": 87.7,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/eva02_e14/ep.txt",
+   "params_incl_bn_stats": 5065193,
+   "file": "eva02_e14/ep_head.pth",
+   "size_mb": 20.3
+  },
+  {
+   "method": "MetaCLIP2",
+   "arch": "ViT-bigG/14",
+   "pretrain": "MetaCLIP2-worldwide (300+ langs)",
+   "image_size": 224,
+   "model": "ViT-bigG-14-worldwide",
+   "loader": "openclip",
+   "pretrain_tag": "metaclip2_worldwide",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 87.11,
+   "table_best_acc1": 87.11,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/metaclip2_bigg/ep.txt",
+   "params_incl_bn_stats": 4490473,
+   "file": "metaclip2_bigg/ep_head.pth",
+   "size_mb": 18.0
+  },
+  {
+   "method": "DINOv3",
+   "arch": "ViT-7B/16",
+   "pretrain": "LVD-1689M",
+   "image_size": 224,
+   "model": "dinov3_vit7b16",
+   "loader": "dinov3",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 88.36,
+   "table_best_acc1": 88.36,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/dinov3_vit7b/ep.txt",
+   "params_incl_bn_stats": 21013481,
+   "file": "dinov3_vit7b/ep_head.pth",
+   "size_mb": 84.1
+  },
+  {
+   "method": "MaskFeat",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.maskfeat/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 71.68,
+   "table_best_acc1": 71.76,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/maskfeat_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "maskfeat_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "I-JEPA",
+   "arch": "ViT-H/14",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_huge_patch14",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_huge_patch14_224.ijepa/checkpoint-799.pth",
+   "extra_flags": "--no_cls_token",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 78.8,
+   "table_best_acc1": 79.0,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/ijepa_vith/ep.txt",
+   "params_incl_bn_stats": 2962921,
+   "file": "ijepa_vith/ep_head.pth",
+   "size_mb": 11.9
+  },
+  {
+   "method": "MoCov3",
+   "arch": "ViT-B/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_base_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_base_patch16_224.mocov3/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 76.21,
+   "table_best_acc1": 76.53,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/mocov3_vitb/ep.txt",
+   "params_incl_bn_stats": 1384937,
+   "file": "mocov3_vitb/ep_head.pth",
+   "size_mb": 5.5
+  },
+  {
+   "method": "iBOT",
+   "arch": "ViT-L/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_large_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_large_patch16_224.ibot/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 79.43,
+   "table_best_acc1": 79.97,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/ibot_vitl/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "ibot_vitl/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "MetaCLIP2",
+   "arch": "ViT-bigG/14-378",
+   "pretrain": "MetaCLIP2-worldwide (300+ langs)",
+   "image_size": 378,
+   "model": "ViT-bigG-14-worldwide-378",
+   "loader": "openclip",
+   "pretrain_tag": "metaclip2_worldwide",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 88.12,
+   "table_best_acc1": 88.12,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/metaclip2_bigg378/ep.txt",
+   "params_incl_bn_stats": 4490473,
+   "file": "metaclip2_bigg378/ep_head.pth",
+   "size_mb": 18.0
+  },
+  {
+   "method": "MaskFeat",
+   "arch": "ViT-L/16",
+   "pretrain": "IN-1K",
+   "image_size": 224,
+   "model": "vit_large_patch16",
+   "loader": "",
+   "pretrain_tag": "",
+   "finetune": "vit_large_patch16_224.maskfeat/checkpoint-799.pth",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 89,
+   "val_acc1_at_head_epoch": 69.56,
+   "table_best_acc1": 69.63,
+   "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+   "training_log": "logs/maskfeat_large/ep.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "maskfeat_large/ep_head.pth",
+   "size_mb": 8.4
+  },
+  {
+   "method": "EVA02-CLIP",
+   "arch": "E-14-plus",
+   "pretrain": "LAION-2B (9B seen)",
+   "image_size": 224,
+   "model": "EVA02-E-14-plus",
+   "loader": "openclip",
+   "pretrain_tag": "laion2b_s9b_b144k",
+   "finetune": "",
+   "extra_flags": "",
+   "cls_features": "ep",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 6,
+   "val_acc1_at_head_epoch": 87.98,
+   "table_best_acc1": 87.98,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/eva02_e14_plus/ep.txt",
+   "params_incl_bn_stats": 5065193,
+   "file": "eva02_e14_plus/ep_head.pth",
+   "size_mb": 20.3
+  },
+  {
+   "method": "Franca",
+   "arch": "ViT-L/14",
+   "pretrain": "LAION-600M",
+   "image_size": 224,
+   "model": "franca_vitl14",
+   "loader": "franca",
+   "pretrain_tag": "",
+   "finetune": "",
+   "extra_flags": "--franca_weights LAION",
+   "cls_features": "ep_all",
+   "ep_queries": 32,
+   "d_out": 1,
+   "head_epoch": 14,
+   "val_acc1_at_head_epoch": 84.28,
+   "table_best_acc1": 84.28,
+   "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+   "training_log": "logs/franca_laion/ep_all.txt",
+   "params_incl_bn_stats": 2108393,
+   "file": "franca_laion/ep_head.pth",
+   "size_mb": 8.4
+  }
+ ],
+ "missing": [
+  {
+   "method": "DiT",
+   "arch": "DiT-XL/2",
+   "reason": "no surviving checkpoint: the run's output directory was deleted after its logs were published; re-run required to export a head"
+  }
+ ]
+}
\ No newline at end of file
diff --git a/maskfeat_large/config.json b/maskfeat_large/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..a6763ad412375f933eb57d77585882d7465dc065
--- /dev/null
+++ b/maskfeat_large/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MaskFeat",
+ "arch": "ViT-L/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_large_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_large_patch16_224.maskfeat/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 69.56,
+ "table_best_acc1": 69.63,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/maskfeat_large/ep.txt"
+}
\ No newline at end of file
diff --git a/maskfeat_large/ep_head.pth b/maskfeat_large/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..0cd283cfbeb2060ffad3dbaf03e562c48a048698
--- /dev/null
+++ b/maskfeat_large/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24d3675d9544fd8374a6dcd8945d03a48213d198158462990db4bb9d583c4b0a
+size 8437265
diff --git a/maskfeat_vitb/config.json b/maskfeat_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d700b42f926894b21188498607469d397640270
--- /dev/null
+++ b/maskfeat_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MaskFeat",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.maskfeat/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 71.68,
+ "table_best_acc1": 71.76,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/maskfeat_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/maskfeat_vitb/ep_head.pth b/maskfeat_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..7d9991bebd54e1fa8a8e44f8afa341afccc671e9
--- /dev/null
+++ b/maskfeat_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:571aa8495e97df53ca4b0ab6a77980ea8afc11ebd1d65943d70ae69183cf91fc
+size 5543441
diff --git a/metaclip2_bigg/config.json b/metaclip2_bigg/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..91bb3d98e3c3110be215f9812439e53896e91ded
--- /dev/null
+++ b/metaclip2_bigg/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MetaCLIP2",
+ "arch": "ViT-bigG/14",
+ "pretrain": "MetaCLIP2-worldwide (300+ langs)",
+ "image_size": 224,
+ "model": "ViT-bigG-14-worldwide",
+ "loader": "openclip",
+ "pretrain_tag": "metaclip2_worldwide",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 87.11,
+ "table_best_acc1": 87.11,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/metaclip2_bigg/ep.txt"
+}
\ No newline at end of file
diff --git a/metaclip2_bigg/ep_head.pth b/metaclip2_bigg/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c6ba9bd4b35ac521dca5f389bc1cb09dcdd1d7ea
--- /dev/null
+++ b/metaclip2_bigg/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffff1af5186d00e4c67b3468ba61e0a439132b09c07b59b10bdbeb1f34e02118
+size 17965457
diff --git a/metaclip2_bigg378/config.json b/metaclip2_bigg378/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..28a27017334a154194641143fdd969ea21dbc9b2
--- /dev/null
+++ b/metaclip2_bigg378/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MetaCLIP2",
+ "arch": "ViT-bigG/14-378",
+ "pretrain": "MetaCLIP2-worldwide (300+ langs)",
+ "image_size": 378,
+ "model": "ViT-bigG-14-worldwide-378",
+ "loader": "openclip",
+ "pretrain_tag": "metaclip2_worldwide",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 88.12,
+ "table_best_acc1": 88.12,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/metaclip2_bigg378/ep.txt"
+}
\ No newline at end of file
diff --git a/metaclip2_bigg378/ep_head.pth b/metaclip2_bigg378/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c87e614f1f8dc0888866558b7a04ae70223796c2
--- /dev/null
+++ b/metaclip2_bigg378/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7d5a9933f113dd404d9b1faeba9c62347e7f1e1b07e16a734dd63df8000ad1b
+size 17965457
diff --git a/mocov3_vitb/config.json b/mocov3_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b4965f85b41f50747643a6dda4829701bb15335
--- /dev/null
+++ b/mocov3_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "MoCov3",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.mocov3/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep_all",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 76.21,
+ "table_best_acc1": 76.53,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/mocov3_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/mocov3_vitb/ep_head.pth b/mocov3_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..c7941220b5495f4a35042709148801e219ac064b
--- /dev/null
+++ b/mocov3_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b69f2c3130204595394c81988ce965ec882574c029052d72a1d5d5165492c5e3
+size 5543441
diff --git a/pe_core_l336/config.json b/pe_core_l336/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..18fe410d0b2154f5fcf6be594e67266097874c1b
--- /dev/null
+++ b/pe_core_l336/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "PE-Core",
+ "arch": "L-14/336",
+ "pretrain": "MetaCLIP-curated 5.4B",
+ "image_size": 336,
+ "model": "PE-Core-L-14-336",
+ "loader": "openclip",
+ "pretrain_tag": "meta",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 12,
+ "val_acc1_at_head_epoch": 87.25,
+ "table_best_acc1": 87.25,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/pe_core_l336/ep.txt"
+}
\ No newline at end of file
diff --git a/pe_core_l336/ep_head.pth b/pe_core_l336/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a3f66149a9bc171ba4e1e2da365a1b1ee747e9b0
--- /dev/null
+++ b/pe_core_l336/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a8bbb1afb39116e04457096d3f7b55adb1db068f4352a5b07d15b311e51fd4e4
+size 8437073
diff --git a/radio_v25_b/config.json b/radio_v25_b/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8b3fbd651ef116c1ae5f2708809df99347cebfe
--- /dev/null
+++ b/radio_v25_b/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "RADIO",
+ "arch": "ViT-B/16",
+ "pretrain": "distilled (DFN5B/SigLIP/DINOv2/SAM)",
+ "image_size": 224,
+ "model": "radio_v2.5-b",
+ "loader": "radio",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 80.26,
+ "table_best_acc1": 80.29,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/radio_v25_b/ep.txt"
+}
\ No newline at end of file
diff --git a/radio_v25_b/ep_head.pth b/radio_v25_b/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b7514717e9cbad3f609ae4bc937b9625d6a63b61
--- /dev/null
+++ b/radio_v25_b/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b38800a4b19a23860df6a9b9f025fa95cc50469bf34008d008337a7a724838d8
+size 5543377
diff --git a/radio_v25_l/config.json b/radio_v25_l/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..1cc6fcc0a005b438639b4b5030d463e91a08850f
--- /dev/null
+++ b/radio_v25_l/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "RADIO",
+ "arch": "ViT-L/16",
+ "pretrain": "distilled (DFN5B/SigLIP/DINOv2/SAM)",
+ "image_size": 224,
+ "model": "radio_v2.5-l",
+ "loader": "radio",
+ "pretrain_tag": "",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 83.4,
+ "table_best_acc1": 83.89,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/radio_v25_l/ep.txt"
+}
\ No newline at end of file
diff --git a/radio_v25_l/ep_head.pth b/radio_v25_l/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..853fb15421e691e126dbab5b1a7f63111bed45bc
--- /dev/null
+++ b/radio_v25_l/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d61e20f290d3262a66e9279ceb1ec96856eb579933a4fa215d7b8081ee8680bb
+size 8437201
diff --git a/siglip2_so400m/config.json b/siglip2_so400m/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c19bfb64d17dc0034383e27bdab2d9b572eb451
--- /dev/null
+++ b/siglip2_so400m/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "SigLIP2",
+ "arch": "SO400M/14",
+ "pretrain": "WebLI (10B imgs, 109 langs)",
+ "image_size": 224,
+ "model": "ViT-SO400M-14-SigLIP2",
+ "loader": "openclip",
+ "pretrain_tag": "webli",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 29,
+ "val_acc1_at_head_epoch": 85.64,
+ "table_best_acc1": 87.68,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/siglip2_so400m/ep.txt"
+}
\ No newline at end of file
diff --git a/siglip2_so400m/ep_head.pth b/siglip2_so400m/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..01cffd590a64668f8471b92c62e7b7103423b0a3
--- /dev/null
+++ b/siglip2_so400m/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6215dc7c33f7a960c045c8d3989a1a5590ab816adc511c722a0ad507b5275a68
+size 10080785
diff --git a/siglip2_vitl/config.json b/siglip2_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..66e51b9a3b85c8d66ff7c66f241b54261896cbb2
--- /dev/null
+++ b/siglip2_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "SigLIP2",
+ "arch": "ViT-L/16",
+ "pretrain": "WebLI (10B imgs, 109 langs)",
+ "image_size": 256,
+ "model": "ViT-L-16-SigLIP2-256",
+ "loader": "openclip",
+ "pretrain_tag": "webli",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 87.06,
+ "table_best_acc1": 87.06,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/siglip2_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/siglip2_vitl/ep_head.pth b/siglip2_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..44ae0803684a9c32403516519e09331a0e5b1262
--- /dev/null
+++ b/siglip2_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52413a02aadddf9f183937cf1a11b38e5457cf58d57a3ffd797762af82e2b602
+size 8437137
diff --git a/siglip_vitl/config.json b/siglip_vitl/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..caaefac1f18e847e4968af4c5a1192dd1e7461fb
--- /dev/null
+++ b/siglip_vitl/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "SigLIP",
+ "arch": "ViT-L/16",
+ "pretrain": "WebLI (10B imgs, 109 langs)",
+ "image_size": 256,
+ "model": "ViT-L-16-SigLIP-256",
+ "loader": "openclip",
+ "pretrain_tag": "webli",
+ "finetune": "",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 6,
+ "val_acc1_at_head_epoch": 85.93,
+ "table_best_acc1": 85.93,
+ "note": "best-epoch head: saved at the run's peak, the number the table reports.",
+ "training_log": "logs/siglip_vitl/ep.txt"
+}
\ No newline at end of file
diff --git a/siglip_vitl/ep_head.pth b/siglip_vitl/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..3817558c56a7fb2fb801fb7f8a400d5fd8fcb2aa
--- /dev/null
+++ b/siglip_vitl/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc9512815df47cb41daf0b5abff4d8fe7e8214254b46b832aa4705a74e922d76
+size 8437073
diff --git a/simmim_vitb/config.json b/simmim_vitb/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b6b5f7259ff8cd750db321399ec30d219e9632d
--- /dev/null
+++ b/simmim_vitb/config.json
@@ -0,0 +1,19 @@
+{
+ "method": "SimMIM",
+ "arch": "ViT-B/16",
+ "pretrain": "IN-1K",
+ "image_size": 224,
+ "model": "vit_base_patch16",
+ "loader": "simmim",
+ "pretrain_tag": "",
+ "finetune": "vit_base_patch16_224.simmim/checkpoint-799.pth",
+ "extra_flags": "",
+ "cls_features": "ep",
+ "ep_queries": 32,
+ "d_out": 1,
+ "head_epoch": 89,
+ "val_acc1_at_head_epoch": 64.81,
+ "table_best_acc1": 64.86,
+ "note": "final-epoch head, not best-epoch: checkpoints overwrite per epoch. table_best_acc1 is the peak of the same run; the difference is the post-peak decline of a frozen-feature probe, largest on VLM encoders.",
+ "training_log": "logs/simmim_vitb/ep.txt"
+}
\ No newline at end of file
diff --git a/simmim_vitb/ep_head.pth b/simmim_vitb/ep_head.pth
new file mode 100644
index 0000000000000000000000000000000000000000..862cf04aa38b22b6046e06d794dba27762ed6788
--- /dev/null
+++ b/simmim_vitb/ep_head.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c29965137e5d1fb0ae776510d9b333114a389c0b2504631f7d7a7a1ac740f09a
+size 5543441