README: Paper section; remove arXiv links; move branch details to Releases

- Add a Paper section naming Chen and Zimmermann (2022) as the dataset's paper
- Remove arXiv links (the hub's Papers sidebar listed only secondary papers);
cite those papers in plain text with titles
- Restore the one-line dataset description and the signal browser link
- Replace the branches paragraph at the top with a pointer to all past releases;
move the branch and revision= explanation to the Releases section

README.md

@@ -28,17 +28,20 @@ configs:
 # Open Source Cross-Sectional Asset Pricing (data releases)
 
 Stock-level signals (firm characteristics) and portfolio returns from
-[Chen and Zimmermann (2022), "Open Source Cross-Sectional Asset Pricing"](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3604626),
-produced by the code at [github.com/OpenSourceAP/CrossSection](https://github.com/OpenSourceAP/CrossSection). 
+[Chen and Zimmermann (2022)](#paper), produced by the code at
+[github.com/OpenSourceAP/CrossSection](https://github.com/OpenSourceAP/CrossSection).
 The official data page and FAQ are at [openassetpricing.com](https://www.openassetpricing.com), which
 includes a [browser for signal descriptions](https://openassetpricing.com/SignalDoc-Browser.html).
 
-**Each data release is its own branch of this repo.** The latest release is on branch
-`Data-Release-2026.10`, and the `main` branch holds a copy of it (plus this README), so
-downloads without `revision=` get the latest data.
+**All past releases are available.** This page shows the latest, 2026.10. See
+[Releases](#releases) for older ones, including 2021.03, which matches the published paper.
+
+## Paper
 
+The paper for this dataset is
+[Chen and Zimmermann (2022), "Open Source Cross-Sectional Asset Pricing," *Critical Finance Review* 11(2): 207–264](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3604626).
 The data are licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). If you
-use them, please cite [Chen and Zimmermann (2022)](#citation).
+use them, please cite the paper ([BibTeX](#citation)).
 
 ## Quick download (latest release, 2026.10)
 
@@ -131,8 +134,8 @@ predictive previously).
 
 
 The placebos may still be useful. Many of them seem to have predictive power and true 
-statistical predictability is plentiful, historically ([Chen (2026, JF Insights)](https://arxiv.org/abs/2206.15365);
-[Chen and Dim (2023)](https://arxiv.org/abs/2311.10685)). 
+statistical predictability is plentiful, historically (Chen 2026, "Most Claimed Statistical Findings in Cross-Sectional Return
+Predictability Are Likely True," *JF Insights*; Chen and Dim 2023, "High-Throughput Asset Pricing").
 
 Firm-level placebo signals are in [Placebos.parquet](https://huggingface.co/datasets/OpenSourceAP/CrossSection/blob/Data-Release-2026.10/Firm%20Level%20Characteristics/Placebos.parquet),
 and their portfolio returns are in
@@ -228,8 +231,12 @@ Older releases may have slightly different file structure.
 
 ## Releases
 
+Each data release is its own branch of this repo. The latest release is on branch
+`Data-Release-2026.10`, and the `main` branch holds a copy of it (plus this README), so
+downloads that don't name a branch (no `revision=` in Python) get the latest data.
+
 For another release, replace `Data-Release-2026.10` in the link with that release's branch
-from the table below. 
+from the table below.
 
 | Release | Branch (`revision=`)     | Notes                                                                                                                                                                                  |
 | ------- | ------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |