Lucene search
+L

12 matches found

EUVD
EUVD
added 2026/07/17 4:52 p.m.12 views

EUVD-2026-18522

vLLM: Processing differential in multi-channel audio downmixing enables hidden-input/moderation bypass for audio models...

7.1CVSS5.2AI score0.00267EPSS
SaveExploits0References6
OSV
OSV
added 2026/07/17 4:52 p.m.6 views

GHSA-6C4R-FMH3-7RH8 vLLM: Processing differential in multi-channel audio downmixing enables hidden-input/moderation bypass for audio models

Issue Description Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results in: - Inconsistency between audio heard by humans e.g., through headphones/regular speakers and...

5.9CVSS5.5AI score0.00267EPSS
SaveExploits0References7
RedhatCVE
RedhatCVE
added 2026/04/06 3:24 p.m.8 views

CVE-2026-34760

A flaw was found in Librosa, a software library used by artificial intelligence AI models like vLLM for processing audio. The library's method for converting stereo audio to mono differs from international standards, causing AI models to interpret audio differently than humans. This inconsistency...

5.9CVSS5.8AI score0.00267EPSS
SaveExploits0References7
PyPA
PyPA
added 2026/04/02 8:16 p.m.5 views

PYSEC-2026-2299

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

7.1CVSS6AI score0.00267EPSS
SaveExploits0References4Affected Software1
OSV
OSV
added 2026/04/02 8:16 p.m.8 views

PYSEC-2026-2299

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

7.1CVSS6AI score0.00267EPSS
SaveExploits0References4
NVD
NVD
added 2026/04/02 8:16 p.m.9 views

CVE-2026-34760

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

7.1CVSS0.00267EPSS
SaveExploits0References4
Cvelist
Cvelist
added 2026/04/02 6:59 p.m.26 views

CVE-2026-34760 vLLM: Downmix Implementation Differences as Attack Vectors Against Audio AI Models

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

5.9CVSS0.00267EPSS
SaveExploits0References4
OSV
OSV
added 2026/04/02 6:59 p.m.8 views

CVE-2026-34760 vLLM: Downmix Implementation Differences as Attack Vectors Against Audio AI Models

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

5.9CVSS5.9AI score
SaveExploits0References6
Vulnrichment
Vulnrichment
added 2026/04/02 6:59 p.m.1 views

CVE-2026-34760 vLLM: Downmix Implementation Differences as Attack Vectors Against Audio AI Models

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

5.9CVSS5.8AI score0.00267EPSS
SaveExploits0References4
ATTACKERKB
ATTACKERKB
added 2026/04/02 6:59 p.m.1 views

CVE-2026-34760

vLLM is an inference and serving engine for large language models LLMs. From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing tomono, while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results...

5.9CVSS5.8AI score0.00267EPSS
SaveExploits0References5Affected Software1
CVE
CVE
added 2026/04/02 6:59 p.m.21 views

CVE-2026-34760

Summary: CVE-2026-34760 concerns vLLM’s audio processing path via Librosa. From version 0.5.5 up to before 0.18.0, Librosa used numpy.mean for mono downmix (to_mono), while ITU-R BS.775-4 specifies a weighted downmix. This mismatch creates inconsistency between audio perceived by humans and audio...

7.1CVSS5.8AI score0.00267EPSS
SaveExploits0References4Affected Software1
Positive Technologies
Positive Technologies
added 2026/04/02 12:0 a.m.12 views

PT-2026-29877

Name of the Vulnerable Software and Affected Versions vLLM versions 0.5.5 through 0.17.999 Description vLLM, an inference and serving engine for large language models LLMs, exhibits an inconsistency in audio processing. Versions 0.5.5 through 0.17.999 utilize numpy.mean for mono downmixing via...

7.1CVSS5.2AI score0.00267EPSS
SaveExploits0References13
Rows per page
Query Builder