A mechanism that simultaneously optimizes revenue while preserving content fidelity is proposed
Explore technology by industry and role
Newest additions first. Explore past and current notes by industry and role.
All notes
Newest additions first. Explore past and current notes by industry and role.
Integrates adaptive group quantization, layer-adaptive sparsification, and prefix-tree parallel verification to accelerate inference
Introduces a taxonomy of commercial influence tiers affecting latent variables in generative processes, from product mentions to long-term preference shaping.
A dataset that identifies how factually correct but out-of-scope context misleads LLMs, even when the content is true.
Proposes a reinforcement learning-based method that considers both structural and sequential dependencies to worsen fairness
A systematic taxonomy of 21 embedding-based diversity measures is proposed, grouped into four computational families.
Reformulates ad insertion as conditional response rewriting, decoupling it from the base model.
The JeNo software successfully applies Lilley's Acoustic Analogy theory to predict jet noise with high accuracy.
Rotating microphone rakes enable precise measurement of modal content in engine noise sources.
Temporal adaptation is introduced at three levels: live-ad tokenization, intent modeling, and preference alignment
LAMA introduces a token-level advertising mechanism that directly embeds advertiser influence into the generation process.
A novel AI-based method for automatically measuring bladder volume using 2D ultrasound images is proposed
LLMs analyze execution reports to explicitly identify bottlenecks and root causes, enabling targeted design improvements
DAIST decomposes package structures into block-level subdomains and uses neural operators, enabling reuse across different topologies without retraining
The first systematic benchmark for assessing multimodal injection attacks on mobile agents is introduced.