Pytorch
This hub aggregates every CVE we track for Pytorch, a product in the oss libraries space. Use it to gauge the current risk picture and drill into individual advisories.
31
CVEs tracked
3
Critical
8
High
0
In CISA KEV
Severity distribution
MEDIUM14HIGH8LOW6CRITICAL3
Monthly trend
1
0
0
0
0
7
4
0
0
0
0
12
0
1
0
1
0
1
0
0
0
0
0
0
2024-102026-09
Latest CVEs
The 15 most recently published vulnerabilities affecting Pytorch.
- CVE-2026-4538PyTorch pt2 Loading deserialization5.3
- CVE-2026-24747PyTorch Vulnerable to Remote Code Execution via Untrusted Checkpoint Files8.8
- CVE-2025-63396An issue was discovered in PyTorch v2.5 and v2.7.1. Omission of profiler.stop() can cause torch.profiler.profile (PythonTracer) to crash or hang during finalization, leading to a Denial of Service ...3.3
- CVE-2025-55552pytorch v2.8.0 was discovered to display unexpected behavior when the components torch.rot90 and torch.randn_like are used together.7.5
- CVE-2025-55560An issue in pytorch v2.7.0 can lead to a Denial of Service (DoS) when a PyTorch model consists of torch.Tensor.to_sparse() and torch.Tensor.to_dense() and is compiled by Inductor.7.5
- CVE-2025-55554pytorch v2.8.0 was discovered to contain an integer overflow in the component torch.nan_to_num-.long().5.3
- CVE-2025-55553A syntax error in the component proxy_tensor.py of pytorch v2.7.0 allows attackers to cause a Denial of Service (DoS).7.5
- CVE-2025-55557A Name Error occurs in pytorch v2.7.0 when a PyTorch model consists of torch.cummin and is compiled by Inductor, leading to a Denial of Service (DoS).7.5
- CVE-2025-55558A buffer overflow occurs in pytorch v2.7.0 when a PyTorch model consists of torch.nn.Conv2d, torch.nn.functional.hardshrink, and torch.Tensor.view-torch.mv() and is compiled by Inductor, leading to...7.5
- CVE-2025-55551An issue in the component torch.linalg.lu of pytorch v2.8.0 allows attackers to cause a Denial of Service (DoS) when performing a slice operation.7.5
- CVE-2025-46150In PyTorch before 2.7.0, when torch.compile is used, FractionalMaxPool2d has inconsistent results.5.3
- CVE-2025-46148In PyTorch through 2.6.0, when eager is used, nn.PairwiseDistance(p=2) produces incorrect results.5.3
- CVE-2025-46152In PyTorch before 2.7.0, bitwise_right_shift produces incorrect output for certain out-of-bounds values of the "other" argument.5.3
- CVE-2025-46149In PyTorch before 2.7.0, when inductor is used, nn.Fold has an assertion error.5.3
- CVE-2025-46153PyTorch 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.Dropout...5.3
Product normalization is registry-driven with AI assist and human review. How it works