{"id":"CVE-2020-15196","aliases":["BIT-tensorflow-2020-15196","GHSA-pg59-2f92-5cph","PYSEC-2020-119","PYSEC-2020-276","PYSEC-2020-311"],"url":"https://o3.security/vulnerability/CVE-2020-15196","summary":"Heap buffer overflow in Tensorflow","details":"In Tensorflow version 2.3.0, the `SparseCountSparseOutput` and `RaggedCountSparseOutput` implementations don't validate that the `weights` tensor has the same shape as the data. The check exists for `DenseCountSparseOutput`, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.","published":"2020-09-25T19:15:14.870Z","modified":"2026-08-07T15:11:43.557559Z","cvss":{"score":9.9,"severity":"CRITICAL","vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H"},"epss":null,"cisaKev":null,"exploitsKnown":null,"affectedPackages":[{"ecosystem":"PyPI","name":"tensorflow","fixedVersion":"2.3.1"},{"ecosystem":"PyPI","name":"tensorflow-cpu","fixedVersion":"2.3.1"},{"ecosystem":"PyPI","name":"tensorflow-gpu","fixedVersion":"2.3.1"}],"fix":{"url":"https://github.com/tensorflow/tensorflow/commit/3cbb917b4714766030b28eba9fb41bb97ce9ee02","label":"tensorflow/tensorflow@3cbb917"},"references":[{"type":"ADVISORY","url":"https://github.com/tensorflow/tensorflow/releases/tag/v2.3.1"},{"type":"FIX","url":"https://github.com/tensorflow/tensorflow/commit/3cbb917b4714766030b28eba9fb41bb97ce9ee02"},{"type":"EVIDENCE","url":"https://github.com/tensorflow/tensorflow/security/advisories/GHSA-pg59-2f92-5cph"}],"provenance":{"sources":["OSV.dev","FIRST.org (EPSS)"],"lastVerified":"2026-08-07T15:11:43.557559Z"}}