Align the card with the paper; update licensor

NOTICE

@@ -1,5 +1,5 @@
 TENDRIL most-embed-de
-Copyright 2026 The TENDRIL authors (IntelliProcure project)
+Copyright 2026 Institute for Public Sector Transformation, Bern University of Applied Sciences
 
 This model is an adapted version of malteos/most-embed-de
 (https://huggingface.co/malteos/most-embed-de). The fine-tuning contribution of most-embed-de is

README.md

@@ -40,17 +40,21 @@ L2-normalized and have 2,048 dimensions, and inputs are truncated at 32,768 toke
 
 ## Results
 
-Test set: [IntelliProcure/TENDRIL](https://huggingface.co/datasets/IntelliProcure/TENDRIL), 35 projects. Scores are
-macro-averaged over projects, and 95% intervals come from 1,000 project-level bootstrap resamples.
+Test set: [IntelliProcure/TENDRIL](https://huggingface.co/datasets/IntelliProcure/TENDRIL) (revision `15eef0bc`),
+33 projects. Scores are macro-averaged over projects, and 95% intervals come from 10,000 project-level bootstrap
+resamples.
 
 | Model | nDCG@10 | Recall@10 | Recall@100 | MRR@10 |
 |---|---|---|---|---|
-| BM25 | 0.437 [0.388, 0.485] | 0.533 | 0.920 | 0.476 |
-| most-embed-de (base) | 0.644 [0.593, 0.695] | 0.763 | 0.977 | 0.665 |
-| **TENDRIL most-embed-de** | **0.660 [0.614, 0.710]** | **0.788** | **0.987** | **0.671** |
+| BM25 | 0.421 [0.381, 0.462] | 0.511 | 0.929 | 0.463 |
+| most-embed-de (base) | 0.640 [0.588, 0.692] | 0.760 | 0.980 | 0.664 |
+| **TENDRIL most-embed-de** | **0.660 [0.613, 0.707]** | **0.787** | **0.987** | **0.673** |
 
-The gain over the base model on the test set is small, and the confidence intervals overlap. On the validation
-projects, the fine-tuned model reaches a balanced nDCG@10 (the mean over areas of action and criteria) of 0.840.
+The gain over the base model is small: +0.020 nDCG@10 [−0.003, +0.044] in a two-sided paired bootstrap over projects.
+After a Holm correction over the four fine-tuned models in the paper, it is significant only for precision@10
+(p < 0.05). With an nDCG@10 of 0.660, the model is on par with the best proprietary model evaluated in the paper
+(Voyage 4 Large, 0.649). On the validation projects, the fine-tuned model reaches a balanced nDCG@10 (the mean
+over areas of action and criteria) of 0.840.
 
 ## Training
 
@@ -58,7 +62,10 @@ projects, the fine-tuned model reaches a balanced nDCG@10 (the mean over areas o
   [TENDRIL-Train](https://huggingface.co/datasets/IntelliProcure/TENDRIL-Train) data: a query, one relevant passage
   and three hard negatives from the same project. Area-of-action and criterion examples were balanced, giving 9,000
   of 13,390 available examples. Validation used held-out projects (63 for areas of action, 62 for criteria).
-- **Loss.** TODO(sophie): name the loss function and how positives and hard negatives were selected.
+- **Loss.** InfoNCE over the positive and the three negatives, with the cosine similarity of L2-normalized embeddings
+  and a similarity scale of 20. Positives are passages the judge rated relevant (2); negatives are passages rated
+  irrelevant (0) for the same query, two from the positive's document and one from another document of the same
+  project (the other pool fills in when one has too few).
 - **Hyperparameters.** Learning rate 1e-5 with linear decay and 10% warmup, AdamW without weight decay, effective
   batch size 32 (2 × 16 gradient accumulation), at most 2 epochs, bf16, seed 42. The model was evaluated every
   56 steps, and the checkpoint with the best balanced validation nDCG@10 (step 168) was kept.
@@ -75,9 +82,17 @@ projects, the fine-tuned model reaches a balanced nDCG@10 (the mean over areas o
 
 ## License
 
-Creative Commons Attribution-NonCommercial 4.0 International (CC-BY-NC-4.0), the same as the base model. See
-[LICENSE](LICENSE) and [NOTICE](NOTICE) for the upstream attributions.
+Creative Commons Attribution-NonCommercial 4.0 International (CC-BY-NC-4.0), the same as the base model. The
+underlying Nemotron-3-Embed-1B is licensed under the OpenMDW License v1.1. See [LICENSE](LICENSE) and
+[NOTICE](NOTICE) for the upstream attributions.
 
 ## Citation
 
-TODO: add the paper citation.
+```bibtex
+@misc{rolshoven2026tendril,
+  title  = {{TENDRIL}: Benchmarking Sustainability-Focused Retrieval in Swiss Public Procurement},
+  author = {Rolshoven, Luca and Probst, Sophie and Matoshi, Veton and Ellendorff, Tilia and Meili, Rahel and
+            Hostettler, Sarah and Jann, Jeremy Austin and Jans, Meret and Tschumi, Pascal and St{\"u}rmer, Matthias},
+  year   = {2026}
+}
+```