CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning

Fine-tuning-as-a-service, while commercially successful for Large Language Model LLM providers, exposes models to harmful fine-tuning attacks. As a widely explored defense paradigm against such attacks, unlearning attempts to remove malicious knowledge from LLMs, thereby essentially preventing th...

kernel: net: bridge: data races indata-races in br_handle_frame_finish()

In the Linux kernel, the following vulnerability has been resolved: net: bridge: use DEVSTATSINC syzbot/KCSAN reported data-races in brhandleframefinish 1 This function can run from multiple cpus without mutual exclusion. Adopt SMP safe DEVSTATSINC to update dev-stats fields. Handles updates to...