Cs-Cr

Do Prompts Guarantee Safety? Mitigating Toxicity from LLM Generations through Subspace Intervention

Cryptography and Security 6 JAN, 2026

Do Prompts Guarantee Safety? Mitigating Toxicity from LLM Generations through Subspace Intervention

By Himanshu Singh

Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks

Artificial Intelligence 6 JAN, 2026

Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks

By Sizhe Chen

Efficient LLM Moderation with Multi-Layer Latent Prototypes

Artificial Intelligence 6 JAN, 2026

Efficient LLM Moderation with Multi-Layer Latent Prototypes

By Maciej Chrabąszcz

SafeCOMM: A Study on Safety Degradation in Fine-Tuned Telecom Large Language Models

Computers and Society 6 JAN, 2026

SafeCOMM: A Study on Safety Degradation in Fine-Tuned Telecom Large Language Models

By Aladin Djuhera

Know Your Scientist: KYC as Biosecurity Infrastructure

Computers and Society 5 JAN, 2026

Know Your Scientist: KYC as Biosecurity Infrastructure

By Jonathan Feldman

Private Sum Computation: Trade-Offs between Communication, Randomness, and Privacy

Cryptography and Security 5 JAN, 2026

Private Sum Computation: Trade-Offs between Communication, Randomness, and Privacy

By Remi A. Chou

When Speculation Spills Secrets: Side Channels via Speculative Decoding In LLMs

Artificial Intelligence 11 JAN, 2026

When Speculation Spills Secrets: Side Channels via Speculative Decoding In LLMs

By Jiankun Wei

Plato's Form: Toward Backdoor Defense-as-a-Service for LLMs with Prototype Representations

Cryptography and Security 6 JAN, 2026

Plato's Form: Toward Backdoor Defense-as-a-Service for LLMs with Prototype Representations

By Chen Chen