Tag: adversarial machine learning
Browse our exclusive articles!
Prevent Unauthorized Distillation of Language Models
Learn how trace rewriting protects large language models from unauthorized knowledge distillation while embedding verifiable watermarks.
AscendKernelGen: LLM-Based Kernel Generation for NPUs
Discover AscendKernelGen, an advanced LLM framework improving kernel generation for Neural Processing Units with 95.5% compilation success.
Persona-Assigned LLMs Show Human-Like Motivated Reasoning
Study reveals persona-assigned large language models exhibit biased reasoning similar to humans, impacting AI reliability in critical societal issues.
Adaptive Dynamic Sampling for Enhanced Math Reasoning AI
Discover SAI-DPO, a dynamic sampling method that adapts training data to model ability, boosting math reasoning performance efficiently.
AI Agents Facing Complex Multi-Objective Decisions
Explore key challenges AI agents face in hard choices, including identification and resolution problems in multi-objective decision-making.