Bug 2398183 (CVE-2025-46153) - CVE-2025-46153 torch: PyTorch inconsistent results
Summary: CVE-2025-46153 torch: PyTorch inconsistent results
Keywords:
Status: NEW
Alias: CVE-2025-46153
Product: Security Response
Classification: Other
Component: vulnerability
Version: unspecified
Hardware: All
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Product Security
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2025-09-25 15:02 UTC by OSIDB Bzimport
Modified: 2025-09-26 18:19 UTC (History)
38 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed:
Embargoed:

Attachments (Terms of Use)

Description OSIDB Bzimport 2025-09-25 15:02:45 UTC 
PyTorch before 3.7.0 has a bernoulli_p decompose function in decompositions.py even though it lacks full consistency with the eager CPU implementation, negatively affecting nn.Dropout1d, nn.Dropout2d, and nn.Dropout3d for fallback_random=True.
Note You need to log in before you can comment on or make changes to this bug.