EntropicaPublic

March updates I need to move some work over from the private repos. The exploration into using massivly multheaded models with mixer and readout functions separated continues. Currently training up a Chinese TinyStories on a 3B-equivalent model that trains, easily, on an 8GB gpu. We're looking at...

CVE-2025-14174-analysis

CVE-2025-14174 Analysis: ANGLE Metal Staging Buffer Out-of-Bounds Write Technical analysis and proof-of-concept for CVE-2025-14174 | ---|--- CVE| CVE-2025-14174 Severity| High Exploited ITW| Yes - targeted attacks on iOS 26 Affected| iOS Safari, macOS Chrome/Chromium/Electron not macOS Safari...

PYSEC-2026-4178 vLLM: Request-selected PyNvVideoCodec GPU decode bypasses static VRAM reservation

Summary Current vLLM main lets an inference request choose the PyNvVideoCodec GPU video decoder through mediaiokwargs.video.videobackend, but engine GPU memory reservation is computed only from static startup configuration and VLLMVIDEOLOADERBACKEND. If the server starts with the default...

PT-2026-105921

Summary Current vLLM main lets an inference request choose the PyNvVideoCodec GPU video decoder through media io kwargs.video.video backend, but engine GPU memory reservation is computed only from static startup configuration and VLLM VIDEO LOADER BACKEND. If the server starts with the default...

CVE-2026-47510

NVIDIA GPU Display Driver for Windows and Linux contains a vulnerability in the kernel mode layer, where a user could cause an integer overflow leading to an out-of-bounds write to GPU memory. A successful exploit of this vulnerability might lead to code execution, denial of service, escalation o...

Allocation of Resources Without Limits or Throttling

Overview vllm is an A high-throughput and memory-efficient inference and serving engine for LLMs Affected versions of this package are vulnerable to Allocation of Resources Without Limits or Throttling via PyNvVideoCodecVideoBackendMixin, where sampler subclass shadowing causes independent...

CVE-2026-100649

A flaw was found in vLLM. An unauthenticated remote attacker can cause a Denial of Service DoS by submitting video processing requests that specify different sampler subclasses. Because active decoder counters are tracked independently across subclasses rather than globally, requests can bypass...

CVE-2026-100649

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

CVE-2026-100649 vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

CVE-2026-100649

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

CVE-2026-100649 vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

EUVD-2026-87743

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

CVE-2026-100649 vLLM before 0.29.0 Resource Limit Bypass via Sampler Subclass

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

CVE-2026-100649

Versions of vLLM prior to 0.29.0 are vulnerable to a resource-limit bypass within the PyNvVideoCodec decoder allocation. The root cause is sampler subclass shadowing , which allows for independent counter increments rather than global tracking. An unauthenticated remote attacker can exploit this ...

CVE-2026-100649: Allocation of Resources Without Limits or Throttling

vLLM before 0.29.0 contains a resource-limit bypass vulnerability in PyNvVideoCodec decoder allocation where sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select different sampler subclasses in video requests to exceed configured decoder limits an...

PT-2026-99320

Name of the Vulnerable Software and Affected Versions vLLM versions prior to 0.29.0 Description A resource-limit bypass exists in the PyNvVideoCodec decoder allocation. This issue occurs because sampler subclass shadowing allows independent counter increments. Unauthenticated attackers can select...

SUSE CVE-2026-94627

vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts,...

CVE-2026-94627

A flaw was found in vLLM. A remote attacker can exploit a vulnerability in the Mooncake connector's management of GPU Key-Value KV cache block ownership. By submitting completion requests with multiple prompts that share a single transfer ID, attackers can cause orphaned KV cache blocks to...

EUVD-2026-84221

vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts,...

CVE-2026-94627

vLLM Mooncake connector through 0.29.0 fails to properly manage GPU KV cache block ownership when concurrent child requests share a single transfer ID in prefill/decode disaggregated deployments. Attackers can trigger GPU memory exhaustion by submitting completion requests with multiple prompts,...