mamba language model framework vulnerable to insecure deserialization when loading pre-trained models from HuggingFace Hub

The mamba language model framework thru 2.2.6 is vulnerable to insecure deserialization CWE-502 when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.frompretrained method uses torch.load to load the pytorchmodel.bin weight file without enabling the security-restrictive...

CVE-2026-31239

The mamba language model framework thru 2.2.6 is vulnerable to insecure deserialization CWE-502 when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.frompretrained method uses torch.load to load the pytorchmodel.bin weight file without enabling the security-restrictive...

Mamba 安全漏洞

Mamba is a state-space model for linear time series modeling, open-sourced by State-Spaces. Versions of Mamba 2.2.6 and earlier contained security vulnerabilities. These vulnerabilities stemmed from the MambaLMHeadModel.frompretrained method, which used torch.load to load weight files without...

CVE-2026-31239

The CVE-2026-31239 entry concerns the Mamba language model framework up to version 2.2.6. The issue is insecure deserialization (CWE-502) when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.from_pretrained() method uses torch.load() to load the pytorch_model.bin weight file...

PT-2026-40126

Name of the Vulnerable Software and Affected Versions mamba versions prior to 2.2.7 Description Insecure deserialization occurs when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.from pretrained function uses torch.load to load the pytorch model.bin weight file without...

CVE-2026-31239

The mamba language model framework thru 2.2.6 is vulnerable to insecure deserialization CWE-502 when loading pre-trained models from HuggingFace Hub. The MambaLMHeadModel.frompretrained method uses torch.load to load the pytorchmodel.bin weight file without enabling the security-restrictive...

Arbitrary Code Injection

Overview diffusers is a State-of-the-art diffusion in PyTorch and JAX. Affected versions of this package are vulnerable to Arbitrary Code Injection in the frompretrained fucntion when a repository contains a None.py file and the custompipeline argument is not supplied. An attacker can execute...

Diffusers has a `trust_remote_code` bypass via `custom_pipeline` and local custom components

Background This vulnerability is found in the DiffusionPipeline.frompretrained flow, which is used to load a pipeline from the HuggingFace Hub. This function accepts an optional custompipeline keyword argument: the name of a Python file in the repo that contains a custom class inheriting from...

Automated Malware Family Classification Using Weighted Hierarchical Ensembles of Large Language Models

Malware family classification remains a challenging task in automated malware analysis, particularly in real-world settings characterized by obfuscation, packing, and rapidly evolving threats. Existing machine learning and deep learning approaches typically depend on labeled datasets, handcrafted...

`trust_remote_code=False` Bypass in LightGlue Nested Config Resolution (Transformers 5.2.0) Leading to Remote Code Execution During Normal `from_pretrained()` Loading

Description Transformers contains a trust-boundary flaw in the LightGlue loading path. When loading a LightGlue model, LightGlueConfig reads trustremotecode from untrusted model config.json and reuses it for nested AutoConfig.frompretrained... resolution. This allows an attacker-controlled model...

Arbitrary Remote Code Execution via `_attn_implementation_internal` Config Injection in transformers (No `trust_remote_code` Required)

Description A critical remote code execution vulnerability exists in the HuggingFace transformers library. An attacker can craft a malicious config.json containing the field attnimplementationinternal set to an attacker-controlled HuggingFace Hub repository ID. When a victim loads this model usin...

cosmos-predict2 (>=1.0.6 <=1.0.9), frankenstein-model (>=5.1.6 <=5.3.9) +11 more potentially affected by CVE-2026-24149 via megatron-core (>=0.10.0 <=0.13.1)

megatron-core PYPI version =0.10.0, =1.0.6, =5.1.6, =0.4.0, =1.0.0, =2.0.8, =2.0.8, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =2.0.5, =5.0.4 Source cves: CVE-2026-24149 Source advisory: SNYK:PYTHON-MEGATRONCORE-15248398...

CVE-2025-13713 Tencent Hunyuan3D-1 load_pretrained Deserialization of Untrusted Data Remote Code Execution Vulnerability

Tencent Hunyuan3D-1 loadpretrained Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Tencent Hunyuan3D-1. User interaction is required to exploit this vulnerability in that the...

CVE-2025-13713 Tencent Hunyuan3D-1 load_pretrained Deserialization of Untrusted Data Remote Code Execution Vulnerability

Tencent Hunyuan3D-1 loadpretrained Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Tencent Hunyuan3D-1. User interaction is required to exploit this vulnerability in that the...

One Leak Away: How Pretrained Model Exposure Amplifies Jailbreak Risks in Finetuned LLMs

Finetuning pretrained large language models LLMs has become the standard paradigm for developing downstream applications. However, its security implications remain unclear, particularly regarding whether finetuned LLMs inherit jailbreak vulnerabilities from their pretrained sources. We investigat...

Defining Cost Function of Steganography with Large Language Models

In this paper, we make the first attempt towards defining cost function of steganography with large language models LLMs, which is totally different from previous works that rely heavily on expert knowledge or require large-scale datasets for cost learning. To achieve this goal, a two-stage...

Tencent Hunyuan3D-1 load_pretrained Deserialization of Untrusted Data Remote Code Execution Vulnerability

This vulnerability allows remote attackers to execute arbitrary code on affected installations of Tencent Hunyuan3D-1. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file. The specific flaw exists within the...

PT-2025-48588

Name of the Vulnerable Software and Affected Versions Tencent Hunyuan3D-1 affected versions not specified Description A flaw exists within the load pretrained function that allows remote attackers to execute arbitrary code on affected installations of Tencent Hunyuan3D-1. The issue is due to...

cosmos-predict2 (>=1.0.6 <=1.0.9), frankenstein-model (>=5.1.6 <=5.3.9) +11 more potentially affected by CVE-2025-23357 via megatron-core (>=0.10.0 <=0.13.1)

megatron-core PYPI version =0.10.0, =1.0.6, =5.1.6, =0.4.0, =1.0.0, =2.0.8, =2.0.8, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =1.0.0, =2.0.5, =5.0.4 Source cves: CVE-2025-23357 Source advisory: SNYK:PYTHON-MEGATRONCORE-13901364...

Backdoor Attacks against Speech Language Models

Large Language Models LLMs and their multimodal extensions are becoming increasingly popular. One common approach to enable multimodality is to cascade domain-specific encoders with an LLM, making the resulting model inherit vulnerabilities from all of its components. In this work, we present the...