https://vulnerability.circl.lu Contains only the most 10 recent entries. http://www.rssboard.org/rss-specification python-feedgen en Thu, 01 Oct 2026 02:06:55 +0000 https://vulnerability.circl.lu/vuln/fkie_cve-2026-93989 <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> https://vulnerability.circl.lu/vuln/fkie_cve-2026-93989 https://vulnerability.circl.lu/vuln/ghsa-pc5q-2x89-w436 <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> https://vulnerability.circl.lu/vuln/ghsa-pc5q-2x89-w436 https://vulnerability.circl.lu/vuln/opensuse-su-2026:11867-1 <p>python313-vllm-0.30.0-1.1 on GA media</p> <p>python313-vllm-0.30.0-1.1 on GA media</p> https://vulnerability.circl.lu/vuln/opensuse-su-2026:11867-1 https://vulnerability.circl.lu/vuln/pysec-2026-4000 <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM through 0.29.0 fails to properly validate bad_words token indices against the model&#39;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.</p> https://vulnerability.circl.lu/vuln/pysec-2026-4000