svm_predict_values in svm.cpp in Libsvm v324, as used in scikit-learn 0.23.2 and other products, allows attackers to cause a denial of service (segmentation fault) via a crafted model SVM (introduced via pickle, json, or any other model permanence standard) with a large value in the _n_support array.
NOTE: the scikit-learn vendor's position is that the behavior can only occur if the library's API is violated by an application that changes a private attribute.

References

• https://nvd.nist.gov/vuln/detail/CVE-2020-28975
• scikit-learn/scikit-learn#18891
• https://github.com/cjlin1/libsvm/blob/9a3a9708926dec87d382c43b203f2ca19c2d56a0/svm.cpp#L2501
• http://packetstormsecurity.com/files/160281/SciKit-Learn-0.23.2-Denial-Of-Service.html
• http://seclists.org/fulldisclosure/2020/Nov/44
• scikit-learn/scikit-learn@1bf13d5
• https://security.gentoo.org/glsa/202301-03
• https://github.com/pypa/advisory-database/tree/main/vulns/scikit-learn/PYSEC-2020-108.yaml