Signal is active with 2 mentions in latest observed window
Recommended action window: Monitor and triage in normal cycle
NVD description
vLLM is an inference and serving engine for large language models. Prior to 0.27.0, an integer overflow in blockIdx.x * 2 * d in activation_kernels.cu can cause act_and_mul_kernel to consume another batched user's input, allowing a request processed in the same inference batch to receive a partial or complete copy of another user's inference result. This issue is fixed in version 0.27.0.
Priority
LOW
Exploitation
NONE
PoC
YES
Patch
AVAILABLE
Momentum
NONE
If you run products in this scope, you should treat this CVE as relevant to your environment.
CVE-2026-73558 vLLM is an inference and serving engine for large language models. Prior to 0.27.0, an integer overflow in blockIdx.x * 2 * d in activation_kernels.cu can cause act_a… https://www.cve.org/CVERecord?id=CVE-2026-73558
Post summary
The text announces CVE‑2026‑73558, detailing an integer overflow in vLLM’s activation kernels that impacts versions earlier than 0.27.0.
🚨*CVE* CVE-2026-73558 vLLM is an inference and serving engine for large language models. Prior to 0.27.0, an integer overflow in blockIdx.x * 2 * d in activation_kernels.cu can cause act_a… https://www.cve.org/CVERecord?id=CVE-2026-73558 ----- Traducción: CVE-2026-73558 vLL… https://infoflow.cloud`
Post summary
The tweet introduces CVE‑2026‑73558, describing an integer overflow in vLLM’s GPU kernel prior to v0.27.0, but offers no PoC, patch, or evidence of active exploitation.
1 of 1 entries
| Part | Vendor | Product | Version | Target SW | Target HW |
|---|---|---|---|---|---|
| App | vllm | vllm | - | - | - |