Division by zero in TFLite's implementation of `EmbeddingLookup`
Published May 14, 2021
2.0
LOWCVSS 4.0
EPSS 0.20%
Description
TensorFlow is an end-to-end open source platform for machine learning. The implementation of the `EmbeddingLookup` TFLite operator is vulnerable to a division by zero error(https://github.com/tensorflow/tensorflow/blob/e4b29809543b250bc9b19678ec4776299dd569ba/tensorflow/lite/kernels/embedding_lookup.cc#L73-L74). An attacker can craft a model such that the first dimension of the `value` input is 0. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.
Affected products
-
- Version < 2.1.4StatusaffectedConstraints-
- Version >= 2.2.0, < 2.2.3StatusaffectedConstraints-
- Version >= 2.3.0, < 2.3.3StatusaffectedConstraints-
- Version >= 2.4.0, < 2.4.2StatusaffectedConstraints-
- Version
Default status is the baseline for the product, each version can override it (e.g. patched versions marked unaffected).
| Vendor | Product | Default status | Versions | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| TensorFlow | TensorFlow | n/a |
|
- < 2.1.4
- ≥ 2.2.0 · < 2.2.3
- ≥ 2.3.0 · < 2.3.3
- ≥ 2.4.0 · < 2.4.2
No data.
No Red Hat product state for this CVE.
No package ranges for this CVE.
Remediation
No remediation recorded yet.
Metrics
CVSS:4.0/AV:L/AC:L/AT:P/PR:L/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
1 other source (GHSA) ▾
CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L
No CVSS v3.0 score for this CVE.
AV:L/AC:L/Au:N/C:P/I:P/A:P
This CVE is not in the KEV list.
No CISA SSVC assessment for this CVE yet.
Estimated probability of exploitation in the wild in the next 30 days (FIRST EPSS). As of Oct 3, 2026.
Score over time
2021–2026- EPSS v1
- EPSS v5
- EPSS v2
- EPSS v3
- EPSS v4
Percentile over time
- EPSS v1
- EPSS v5
- EPSS v2
- EPSS v3
- EPSS v4
Table of values (10 key points)
Flat stretches are collapsed; showing up to 120 newest points.
| Date | Score | Percentile | Model |
|---|---|---|---|
| Oct 3, 2026 | 0.20% (0.00201) | 9.02th | v5 (v2026.06.15) |
| Jun 15, 2026 | 0.20% (0.00201) | 9.99th | v5 (v2026.06.15) |
| Mar 17, 2025 | 0.01% (0.00014) | 1.20th | v4 (v2025.03.14) |
| Dec 12, 2024 | 0.05% (0.00048) | 19.96th | v3 (v2023.03.01) |
| May 8, 2024 | 0.05% (0.00048) | 16.78th | v3 (v2023.03.01) |
| Mar 7, 2023 | 0.05% (0.00048) | 14.74th | v3 (v2023.03.01) |
| Mar 6, 2023 | 0.89% (0.00885) | 27.89th | v2 (v2022.01.01) |
| Feb 4, 2022 | 1.40% (0.01404) | 49.02th | v2 (v2022.01.01) |
| Feb 3, 2022 | 1.15% (0.01146) | 29.80th | v5 (v2026.06.15) |
| May 15, 2021 | 0.62% (0.00624) | 0.00th | v1 |
References (8)
- https://github.com/advisories/GHSA-4vrf-ff7v-hpgr Advisory
- https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2021-524.yaml
- https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2021-722.yaml
- https://github.com/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2021-233.yaml
- https://github.com/tensorflow/tensorflow/blob/e4b29809543b250bc9b19678ec4776299dd569ba/tensorflow/lite/kernels/embedding_lookup.cc#L73-L74
- https://github.com/tensorflow/tensorflow/commit/f61c57bd425878be108ec787f4d96390579fb83e x_refsource_MISCPatchThird Party Advisory
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-4vrf-ff7v-hpgr x_refsource_CONFIRMExploitPatchThird Party Advisory
- https://nvd.nist.gov/vuln/detail/CVE-2021-29596
Change history (0)
No recorded changes yet.