NVIDIA
US Launches Review of Nvidia AI Chip Sales to China
The Biden administration initiates a new review of Nvidia's advanced AI chip exports to China, potentially tightening restrictions on technology critical for training generative AI models.
Latest
mechanistic interpretability
SALVE: New Technique Enables Mechanistic Control of Neural Networ
Researchers introduce SALVE, combining sparse autoencoders with latent vector editing for precise mechanistic control over neural network behaviors and outputs.
AMD
AMD's AIE4ML Framework: Compiling Neural Networks for NPUs
AMD researchers unveil AIE4ML, an end-to-end compiler framework that maps neural networks to next-gen AI Engines, achieving significant speedups over CPU implementations for ML workloads.
LLM
New Method Speeds Up LLM Data Valuation with Shapley Values
Researchers propose efficient Shapley value approximation using language model arithmetic to determine which training data samples matter most for LLM fine-tuning.
multi-modal AI
MMGR: New Framework Unifies Multi-Modal Generative Reasoning
New research introduces MMGR, a framework that enables AI models to perform generative reasoning across multiple modalities including text, images, and video.
deepfake detection
MITRE Warns of Deepfake KYC Attacks Using Face-Swap Tools
MITRE has identified deepfake face-swap tools as a growing threat to Know Your Customer identity verification systems, highlighting vulnerabilities in financial authentication.
deepfake
Coalition Launches Deepfake Coverage in Cyber Insurance
Cyber insurance provider Coalition adds deepfake-specific coverage to policies, signaling growing recognition of synthetic media fraud risks in enterprise security.
LLM
LLMs as Neural Architects: Auto-Generating Image Captioning Model
New research demonstrates LLMs can design complete neural network architectures for image captioning under strict API constraints, opening new possibilities for automated AI system design.
AI Security
API Access Enables AI Model Cloning and Safety Bypass
New research reveals how anyone with API access can clone AI models and strip away safety guardrails, creating unregulated copies capable of generating harmful content.
embeddings
Why Embeddings Make LLMs Seem Intelligent: A Technical Primer
The famous equation 'King - Man + Woman = Queen' reveals how embeddings capture semantic meaning in vector space, forming the foundation of why large language models appear intelligent.
AI Security
Browser Extensions Harvesting AI Chats From 8M Users
Security researchers discover browser extensions with 8 million users secretly collecting extended conversations from ChatGPT, Gemini, and other AI platforms, raising major privacy concerns.
AI Agents
Patronus AI Tackles 63% Agent Failure Rate With Living Worlds
New benchmark reveals AI agents fail 63% of complex tasks. Patronus AI's dynamic simulation environments aim to fix reliability crisis plaguing autonomous systems.