diff --git a/neurips/paper_architecture_of_errors/submission/references.bib b/neurips/paper_architecture_of_errors/submission/references.bib
index 2789f8f..7aec529 100644
--- a/neurips/paper_architecture_of_errors/submission/references.bib
+++ b/neurips/paper_architecture_of_errors/submission/references.bib
@@ -39,13 +39,10 @@ @article{zakka2024almanac
 }
 
 @article{arbuzov2025beyond,
-  author       = {{Anonymous}},
-  title        = {Beyond Exponential Decay: Rethinking Error Accumulation in Large Language Models},
-  journal      = {arXiv preprint},
-  year         = {2025},
-  eprint       = {2505.24187},
-  archivePrefix = {arXiv},
-  note         = {Author field anonymized for double-blind submission; arXiv ID public.},
+  title={Beyond exponential decay: Rethinking error accumulation in large language models},
+  author={Arbuzov, Mikhail L and Shvets, Alexey A and Beir, Sisong},
+  journal={arXiv preprint arXiv:2505.24187},
+  year={2025}
 }
 
 @inproceedings{asai2024selfrag,