Add40epoch DNA-only SSEL weights, inference settings and validation summary

README.md

@@ -11,7 +11,7 @@ tags:
 
 # Model Card for SSEL
 
-Checkpoints and extracted DNA embeddings for **Hierarchical Embedding of DNA Barcode Taxonomy with Hyperbolic Entailment Cones**. This repository contains the nine 768D experiments presented in the iBOL poster; eight also appear in the UniReps submission.
+Checkpoints and extracted DNA embeddings for **Hierarchical Embedding of DNA Barcode Taxonomy with Hyperbolic Entailment Cones**. This repository contains the nine 768D experiments presented in the iBOL poster (eight also appear in the UniReps submission), plus an additional 40-epoch DNA-only SSEL checkpoint.
 
 ## Model Details
 
@@ -34,18 +34,19 @@ The DNA branch starts from [BarcodeBERT](https://huggingface.co/bioscan-ml/clibd
 | EL | [el](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_EL_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_EL_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_EL_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 | SEL | [sel](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_basic_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_no_radius_barrier_4rank_posneg_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 | Positive EL + CL | [positive-el-cl](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_MERU_style_4rank_posEL02_CL1_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_MERU_style_4rank_posEL02_CL1_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_MERU_style_4rank_posEL02_CL1_768d_nonres_gap05_15epochs_seed42/config.yaml) | No |
-| SSEL | [ssel](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/config.yaml) | Yes |
+| SSEL (40 epochs) | [ssel-40epochs](Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/last.pth) | — | [Inference config](Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/inference_config.json) | No (additional checkpoint) |
+| SSEL (15 epochs, UniReps) | [ssel](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/SEL_structure_v2_768d_dna_nonres_margin05_gap05_weight10_15epochs_seed42/config.yaml) | Yes |
 | No direct placement | [no-direct-placement](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_full_no_direct_placement_768d_nonres_gap05_15epochs_seed42_retry_20261008/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_full_no_direct_placement_768d_nonres_gap05_15epochs_seed42_retry_20261008/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_full_no_direct_placement_768d_nonres_gap05_15epochs_seed42_retry_20261008/config.yaml) | Yes |
 | Frozen labels | [frozen-labels](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_full_frozen_labels_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_full_frozen_labels_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_full_frozen_labels_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 | SEL + nesting | [sel-nesting](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_nesting_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_basic_plus_nesting_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_nesting_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 | SEL + separation | [sel-separation](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_sibling_separation_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_basic_plus_sibling_separation_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_sibling_separation_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 | SEL + allocation | [sel-allocation](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_allocation_768d_nonres_gap05_15epochs_seed42/last.pth) | [DNA](Ver_1_0/extracted_embedding/bioscan_5m/UniReps_SEL_basic_plus_allocation_768d_nonres_gap05_15epochs_seed42/extracted_feature_from_test_split.hdf5) | [Config](Ver_1_0/ckpt/bioscan_clip/bioscan_5m/UniReps_SEL_basic_plus_allocation_768d_nonres_gap05_15epochs_seed42/config.yaml) | Yes |
 
-Each selected checkpoint is the exact **last.pth** used for its reported result. All nine runs use 768 dimensions, 15 epochs and seed 42. EL and SEL omit minimum distance; the other listed variants include it. Positive EL + CL is our DNA–taxonomy adaptation, not an exact reproduction of image–text MERU.
+Each selected checkpoint is the exact **last.pth** used for its reported result. The nine original `Ver_1_0` runs use 768 dimensions, 15 epochs and seed 42. The additional `ssel-40epochs` checkpoint uses 768 dimensions, 40 epochs and seed 42; its [full validation summary](Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/validation_summary.json) reports 99.56% seen Species nearest-label accuracy, 82.78% unseen-species detection and 2.52% seen false rejection on 14,757 seen and 8,819 unseen validation queries. It is separate from the original 15-epoch UniReps checkpoint. EL and SEL omit minimum distance; the other listed variants include it. Positive EL + CL is our DNA–taxonomy adaptation, not an exact reproduction of image–text MERU.
 
 ## Files and Usage
 
-All versioned artifacts are under **Ver_1_0/**. The layout follows CLIBD's `ckpt/` and `extracted_embedding/` structure, with the version directory placed at the outside:
+The original poster artifacts are under **Ver_1_0/**. The added 40-epoch checkpoint and its inference-only configuration are under **Ver_1_1/**; no test embeddings are published for this checkpoint. The layout follows CLIBD's `ckpt/` and `extracted_embedding/` structure, with the version directory placed at the outside:
 
 ```text
 Ver_1_0/

Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/inference_config.json

@@ -0,0 +1,47 @@
+{
+  "source_config_sha256": "87376a228c7048a9d763f5116f2908eb278de2ec07a0f30080cb7826c42d7825",
+  "config": {
+    "model_config": {
+      "model_output_name": "SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42",
+      "dna_output_pipeline": "transformed_hidden_state_mean",
+      "output_dim": 768,
+      "for_bio_clip": false,
+      "dna": {
+        "model": "barcode_bert",
+        "input_type": "sequence"
+      },
+      "hyperbolic_space": {
+        "curv_init": 1.0,
+        "min_radius": 0.1,
+        "sel_clamp_feature_alpha_max": true,
+        "entailment_geometry": "hyperbolic",
+        "nearest_neighbor_metric_cross_modality": "angle",
+        "inference_project_to_hyperbolic_space": true,
+        "stacked_entailment": {
+          "embedding_dim": 768,
+          "label_table": {
+            "enabled": true,
+            "add_dna_bin_level": false
+          },
+          "dna_projection": {
+            "enabled": true,
+            "type": "mlp",
+            "activation": "gelu",
+            "dropout": 0.0,
+            "input_dim": 768,
+            "hidden_dim": 1536,
+            "output_dim": 768,
+            "residual": false,
+            "normalize": false
+          }
+        }
+      }
+    },
+    "barcodebert_setting": {
+      "old_model_setting": {
+        "k": 5,
+        "max_len": 660
+      }
+    }
+  }
+}

Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/last.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25680fbc63925526707d1ff523fa6aa39ae445c1310c30de61aa9db10d87c9bd
+size 452784372

Ver_1_1/ckpt/bioscan_clip/bioscan_5m/SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42/validation_summary.json

@@ -0,0 +1,40 @@
+{
+  "model": "ssel-40epochs",
+  "model_output_name": "SSEL_granularity_comparison_768d_dna_only_nonres_40epochs_seed42",
+  "epochs": 40,
+  "modality": "DNA-only",
+  "dimension": 768,
+  "checkpoint_sha256": "25680fbc63925526707d1ff523fa6aa39ae445c1310c30de61aa9db10d87c9bd",
+  "source_config_sha256": "87376a228c7048a9d763f5116f2908eb278de2ec07a0f30080cb7826c42d7825",
+  "source_job_id": "347623",
+  "evaluation": "full validation; nearest exterior angle and outside all training-known Species cones",
+  "candidate_counts": {
+    "order": 55,
+    "family": 929,
+    "genus": 6819,
+    "species": 11846
+  },
+  "units": "percent; micro",
+  "validation": {
+    "val_seen": {
+      "total": 14757,
+      "rejected": 372,
+      "reject_percent": 2.5208375686115065,
+      "top1_rank": "species",
+      "top1_correct": 14692,
+      "top1_percent": 99.55953107000067,
+      "ordered_processids_sha256": "047de23d9407f8c087e6bdd6453a109c67cdd88e39db3d5913916b12fd8ab66e"
+    },
+    "val_unseen": {
+      "total": 8819,
+      "rejected": 7300,
+      "reject_percent": 82.7758249234607,
+      "top1_rank": "genus",
+      "top1_correct": 7566,
+      "top1_percent": 85.79203991382244,
+      "ordered_processids_sha256": "075b2f9addb2a2fc70708d130d1338bbf395d0af2c2bd1f374564fea22b38bfc"
+    }
+  },
+  "source_evaluation_completed_utc": "2026-11-12T18:22:53.752698+00:00",
+  "note": "Additional40epoch checkpoint; original15epoch UniReps weights are retained. Validation results are not test-set results."
+}