Executive brief
HashNeRF-pytorch is a machine learning tool used for training and rendering neural graphics. A security flaw allows an attacker to execute malicious code on a user's system if the user is tricked into loading a specially crafted 'checkpoint' file (a file used to save training progress). This could lead to the theft of sensitive data, such as private datasets, API keys, or cloud credentials stored on the machine.
Technical details
A deserialization vulnerability exists in the Checkpoint File Handler of HashNeRF-pytorch. The 'run_nerf.py' script utilizes the 'torch.load()' function without the 'weights_only=True' parameter, which defaults to using the 'pickle' module for deserialization. An attacker can exploit this by providing a malicious checkpoint file via the '--ft_path' argument or by placing it in the experiment log directory. Successful exploitation allows for arbitrary code execution in the context of the user running the script. A pull request (PR #50) has been submitted to implement safe loading but is currently awaiting acceptance.
Affected products
- yashbhalgat HashNeRF-pytorch up to 82885e698295982504eb6a26d060a6b2473e3706
Timeline
- 2026-06-10: disclosed: Issue and PR opened on GitHub
- 2026-07-13: advisory: CVE published by NVD/VulDB