vLLM through 0.29.0 Cross-Request Logits Corruption via bad_words
8Vexday Risk Score
No sign of exploitation. No public exploitation artifact known so far.
ssvc Trackcvss 2.3epss 0.2%
exploitation probability
0.2%top 90% of all CVEs
observed exploitation
nono source reports it
vLLM through 0.29.0 fails to properly validate bad_words token indices against the model's generation output width in SamplingParams.update_from_tokenizer(). Attackers can supply out-of-bounds token indices that corrupt logits memory of concurrent requests, causing different in-flight HTTP requests to return incorrect tokens.
CVSS:4.0/AV:N/AC:H/AT:P/PR:L/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N
Affected products
vllm-project · vllmReferences
https://github.com/vllm-project/vllmhttps://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/sampling_params.py#L694-L753https://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/v1/worker/gpu/sample/bad_words.pyhttps://github.com/vllm-project/vllm/pull/48824https://www.vulncheck.com/advisories/vllm-through-0.29.0-cross-request-logits-corruption-via-bad-words