Vulnerability · Published Sep 25, 2025 · Assigned by mitre
CVE-2025-46153
Medium 5.3
In plain English
A medium-severity security flaw in the pytorch package (Debian). There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.4% (higher than 36% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.
Severity (CVSS 3.1)5.3Medium
Exploit odds (30 days)0.4%top 63.6% of all CVEs
Exploited in the wildNo reportsCISA SSVC: none
Attack toolingNoneno public tooling known
Official description
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.
Affected packages & fixed versions
| Ecosystem | Package | Fixed in | Advisory |
|---|---|---|---|
| Debian:12 | pytorch | no fix listed | DEBIAN-CVE-2025-46153 |
| Debian:13 | pytorch | no fix listed | DEBIAN-CVE-2025-46153 |
| Debian:14 | pytorch | 2.12.0+dfsg2-1 | DEBIAN-CVE-2025-46153 |
| PyPI | torch | 2.7.0 | PYSEC-2025-202 |
| Ubuntu:22.04:LTS | pytorch | no fix listed | UBUNTU-CVE-2025-46153 |
| Ubuntu:25.10 | pytorch | no fix listed | UBUNTU-CVE-2025-46153 |
| Ubuntu:26.04:LTS | pytorch | no fix listed | UBUNTU-CVE-2025-46153 |
| Ubuntu:26.10 | pytorch | no fix listed | UBUNTU-CVE-2025-46153 |
Affected products (as reported)
Weakness: CWE-1176 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N