Update dataset card with paper link, GitHub repository, and citation

README.md

@@ -1,14 +1,13 @@
 ---
-license: cc-by-4.0
 language:
 - en
-pretty_name: CGM-JEPA Pretraining Corpus
+license: cc-by-4.0
+size_categories:
+- 100K<n<1M
 task_categories:
+- time-series-forecasting
 - feature-extraction
-- other
-modalities:
-- Time Series
-- Tabular
+pretty_name: CGM-JEPA Pretraining Corpus
 tags:
 - continuous-glucose-monitor
 - cgm
@@ -20,8 +19,6 @@ tags:
 - time-series
 - pretraining
 - jepa
-size_categories:
-- 100K<n<1M
 configs:
 - config_name: default
   data_files:
@@ -31,7 +28,9 @@ configs:
 
 # CGM-JEPA Pretraining Corpus
 
-Continuous glucose monitor (CGM) time-series corpus used for self-supervised pretraining of **CGM-JEPA**, **X-CGM-JEPA**, **GluFormer**, and **TS2Vec** encoders in the paper *CGM-JEPA: Learning Consistent Continuous Glucose Monitor Representations via Predictive Self-Supervised Pretraining*.
+Continuous glucose monitor (CGM) time-series corpus used for self-supervised pretraining of **CGM-JEPA**, **X-CGM-JEPA**, **GluFormer**, and **TS2Vec** encoders in the paper [CGM-JEPA: Learning Consistent Continuous Glucose Monitor Representations via Predictive Self-Supervised Pretraining](https://huggingface.co/papers/2605.00933).
+
+**Code:** [https://github.com/cruiseresearchgroup/CGM-JEPA](https://github.com/cruiseresearchgroup/CGM-JEPA)
 
 > Pretraining-only corpus. For the labeled downstream-evaluation cohorts (insulin resistance and β-cell dysfunction classification), see [`CRUISEResearchGroup/CGM-JEPA-Downstream`](https://huggingface.co/datasets/CRUISEResearchGroup/CGM-JEPA-Downstream). For pretrained model weights, see [`CRUISEResearchGroup/CGM-JEPA`](https://huggingface.co/CRUISEResearchGroup/CGM-JEPA).
 
@@ -129,8 +128,11 @@ When using this dataset, please cite **all three sources**: the Stanford CGM Stu
 
 ## Citation
 
-> _Citation block to be filled once the CGM-JEPA paper has a stable venue / arXiv link._
-
-## Code repository
-
-[`https://github.com/cruiseresearchgroup/CGM-JEPA`](https://github.com/cruiseresearchgroup/CGM-JEPA)
+```bibtex
+@article{muhammad2026cgm,
+  title   = {CGM-JEPA: Learning Consistent Continuous Glucose Monitor Representations via Predictive Self-Supervised Pretraining},
+  author  = {Muhammad, Hada Melino and Li, Zechen and Salim, Flora and Metwally, Ahmed A},
+  journal = {arXiv preprint arXiv:2605.00933},
+  year    = {2026}
+}
+```