And the troops cheering on the lies & fanaticism was beyond “disturbing”! I’m sick of the “ya but” media normalizing this ctrap & moving on!

The workshop is almost over! 🎉 Ending with a tour and live demo of Lumicks and their Ctrap system! #Workshop #Lumicks #Ctrap

Holy ctrap. Is that a Gibson Grabber?

🌟 Learn about the pivotal role of MeCP2 in epigenetic regulation with the latest research from Prof. Shixin Liu’s lab! Explore the dynamic world of MeCP2 in our latest digest! Get the full story: https://buff.ly/40WY5CD

#MeCP2 #RettSyndrome #Epigenetics #SingleMolecule #CTrap

CTRAP is a promising pre-deployment alignment method that makes AI models resistant to harmful fine-tuning by causing them to "break" if malicious tuning occurs, while remaining stable under benign changes.

anonymous.4open.science/r/CTRAP/READ...

🌟 Learn about the pivotal role of MeCP2 in epigenetic regulation with the latest research from Prof. Shixin Liu’s lab! Explore the dynamic world of MeCP2 in our latest digest! Get the full story: https://buff.ly/4298VpL

#MeCP2 #RettSyndrome #Epigenetics #SingleMolecule #CTrap

Isn’t it beautiful how scientists braid DNA at the molecular level? Watch our tutorial to see the C-Trap in action and discover the fascinating process of DNA braiding!

Movie credit: Ian Hickson lab

#DNABraiding #SingleMolecule #LUMICKS #CTrap

Learn about C-Trap: https://buff.ly/4h03X2X

Biao Yi, Tiansheng Huang, Baolei Zhang, Tong Li, Lihai Nie, Zheli Liu, Li Shen
CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning
https://arxiv.org/abs/2505.16559

5 random letters (CTRAP)

counters harmful fine-tuning risks across various LLMs and attack settings, while maintaining high performance in benign scenarios. Our code is available at https://anonymous.4open.science/r/CTRAP. [8/8 of https://arxiv.org/abs/2505.16559v1]

for the attacker. Crucially, this collapse mechanism remains dormant during benign fine-tuning, ensuring the model's utility and general capabilities are preserved for legitimate users. Extensive empirical results demonstrate that CTRAP effectively [7/8 of https://arxiv.org/abs/2505.16559v1]

issue unaddressed by selective unlearning. We introduce the Collapse Trap (CTRAP) as a practical mechanism to implement this concept conditionally. Embedded during alignment, CTRAP pre-configures the model's reaction to subsequent fine-tuning [5/8 of https://arxiv.org/abs/2505.16559v1]

Biao Yi, Tiansheng Huang, Baolei Zhang, Tong Li, Lihai Nie, Zheli Liu, Li Shen: CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning https://arxiv.org/abs/2505.16559 https://arxiv.org/pdf/2505.16559 https://arxiv.org/html/2505.16559

My first music video

youtu.be/EU_kI3mPv_g?...