CVE-2026-53923
MEDIUM
MEDIUM
Priority: 0.03
BB Relevance:
61/100
Published: 2026-06-22
Modified: 2026-10-25
First seen: 2026-10-29
Status: active
Attack Surface
Network
Low Complexity
No Auth
CVSS Vector Breakdown
AV:
Network
AC:
Low
PR:
None
UI:
Passive
S:
X
AT:
None
VC:
Low
VI:
Low
VA:
None
Description
vLLM is an inference and serving engine for large language models (LLMs). From 0.5.5 until 0.23.1rc0, integer truncation of tensor dimensions in vLLM's GGUF dequantize kernels (csrc/quantization/gguf/gguf_kernel.cu) causes partial tensor processing. The output tensor is allocated at full size via torch::empty (uninitialized memory), but the dequantize CUDA kernel processes only a truncated number of elements. The unfilled portion of the output tensor retains whatever was previously in GPU memory. In multi-tenant inference deployments, this residual GPU memory may contain tensor data from other users' inference requests, constituting information disclosure. This vulnerability is fixed in 0.23.1rc0.
References & Advisories (3)
Third Party Advisory
Patch
Issue Tracking