`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...

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...

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...

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...

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...

CVE-2025-23348

NVIDIA Megatron-LM for all platforms contains a vulnerability in the pretraingpt script, where malicious data created by an attacker may cause a code injection issue. A successful exploit of this vulnerability may lead to code execution, escalation of privileges, information disclosure, and data...

GHSA-6VM5-6JV9-RJPJ MONAI: Unsafe torch usage may lead to arbitrary code execution

Summary In modeldict = torch.loadfullpath, maplocation=torch.devicedevice, weightsonly=True in monai/bundle/scripts.py , weightsonly=True is loaded securely. However, insecure loading methods still exist elsewhere in the project, such as when loading checkpoints. This is a common practice when...

Don't Throw the Baby out with the Bathwater: How and Why Deep Learning for ARC

The Abstraction and Reasoning Corpus ARC-AGI presents a formidable challenge for AI systems. Despite the typically low performance on ARC, the deep learning paradigm remains the most effective known strategy for generating skillful state-of-the-art neural networks NN across varied modalities and...

Can In-Context Reinforcement Learning Recover from Reward Poisoning Attacks?

We study the corruption-robustness of in-context reinforcement learning ICRL, focusing on the Decision-Pretrained Transformer DPT, Lee et al., 2023. To address the challenge of reward poisoning attacks targeting the DPT, we propose a novel adversarial training framework, called Adversarially...

Adversarially Robust Spiking Neural Networks with Sparse Connectivity

Deployment of deep neural networks in resource-constrained embedded systems requires innovative algorithmic solutions to facilitate their energy and memory efficiency. To further ensure the reliability of these systems against malicious actors, recent works have extensively studied adversarial...

Secure Transfer Learning: Training Clean Models against Backdoor in (Both) Pre-Trained Encoders and Downstream Datasets

Transfer learning from pre-trained encoders has become essential in modern machine learning, enabling efficient model adaptation across diverse tasks. However, this combination of pre-training and downstream adaptation creates an expanded attack surface, exposing models to sophisticated backdoor...

PT-2024-26627

Name of the Vulnerable Software and Affected Versions huggingface/transformers affected versions not specified Description The issue allows for arbitrary code execution through deserialization of untrusted data within the load repo checkpoint function of the TFPreTrainedModel class. Attackers can...

Deserialization of untrusted data

UNSUPPORTED WHEN ASSIGNED A vulnerability was found in DeepFaceLab pretrained DF.wf.288res.384.92.72.22 and classified as problematic. This issue affects the function applyxseg of the file main.py. The manipulation leads to deserialization. The attack may be initiated remotely. The complexity of ...

CVE-2024-0654

A vulnerability, which was classified as problematic, was found in DeepFaceLab pretrained DF.wf.288res.384.92.72.22. Affected is an unknown function of the file mainscripts/Util.py. The manipulation leads to deserialization. Local access is required to approach this attack. The exploit has been...

Deserialization of untrusted data

A vulnerability, which was classified as problematic, was found in DeepFaceLab pretrained DF.wf.288res.384.92.72.22. Affected is an unknown function of the file mainscripts/Util.py. The manipulation leads to deserialization. Local access is required to approach this attack. The exploit has been...

CVE-2024-0654 DeepFaceLab Util.py deserialization

A vulnerability, which was classified as problematic, was found in DeepFaceLab pretrained DF.wf.288res.384.92.72.22. Affected is an unknown function of the file mainscripts/Util.py. The manipulation leads to deserialization. Local access is required to approach this attack. The exploit has been...