CVE-2026-44223
CVE.org · Published
- Record status
- Published
- Identifier year (not publication year)
- 2026
- Source publication
- Source updated
- Imported by cveboard
- Source snapshot
- cve_2026-10-24_midnight_baseline ·
CNA source text (English)
vLLM: extract_hidden_states speculative decoding crashes server on any request with penalty parameters
vLLM is an inference and serving engine for large language models (LLMs). From 0.18.0 to before 0.20.0, the extract_hidden_states speculative decoding proposer in vLLM returns a tensor with an incorrect shape after the first decode step, causing a RuntimeError that crashes the EngineCore process. The crash is triggered when any request in the batch uses sampling penalty parameters (repetition_penalty, frequency_penalty, or presence_penalty). A single request with a penalty parameter (e.g., "repetition_penalty": 1.1) is sufficient to crash the server. This vulnerability is fixed in 0.20.0.
Scoped product and version facts
Fix information unknown; no fixed version is derived here. Missing information does not prove that no fix exists.
Exploitation unknown; this catalogue does not assess exploitation.
vllm
- Vendor
- vllm-project
- Product
- vllm
- Platforms
- Unknown
- Default status
Unknown
| Version (version) | Status (status) | Version scheme (versionType) | Exclusive end (lessThan) | Inclusive end (lessThanOrEqual) | Status changes (at → status) |
|---|---|---|---|---|---|
| >= 0.18.0, < 0.20.0 | affected | Unavailable | Unavailable | Unavailable | None listed |