LLM Research
Sparse Autoencoders Enable Fine-Grained Control of LLM Reasoning
New research demonstrates how Sparse Autoencoders can steer LLM reasoning processes, enabling precise control over chain-of-thought behavior without retraining models.
Editorial Team
LLM
Multi-Agent System Automates LLM Prompt Optimization
New research introduces an evaluation-driven multi-agent workflow that automatically optimizes prompt instructions for improved LLM instruction following performance.
AI Certification
Embodied AI Certification Framework Proposes Trust Metrics
New research proposes maturity-based certification for embodied AI systems, introducing quantifiable trustworthiness metrics that could reshape how we evaluate AI reliability and authenticity.
deepfake regulation
Italy Privacy Regulator Warns xAI's Grok Over Deepfake Content
Italy's data protection authority issues formal warning to xAI over Grok chatbot's handling of deepfake AI-generated content, signaling increased EU regulatory pressure on synthetic media.
OpenAI
Musk vs OpenAI Lawsuit Heads to Jury Trial in March
Elon Musk's high-stakes legal battle against OpenAI will be decided by jury in March, with implications for AI governance and the future of frontier AI development.
deepfake detection
iProov and HYPR Partner to Fight Deepfake Workforce Fraud
Identity verification leaders iProov and HYPR join forces to combat deepfake-powered workforce fraud, combining biometric authentication with phishing-resistant identity solutions.
AI security
Privacy Attacks Target Graph Diffusion Models: New Security Risks
New research reveals three classes of inference attacks against graph generative diffusion models, exposing membership inference, property inference, and data reconstruction vulnerabilities in AI generation systems.
LLM Security
Zero-Shot LLM Jailbreak Detection via Internal Discrepancy
New research proposes ALERT, a training-free method to detect jailbreak attacks on LLMs by analyzing discrepancies between internal model representations and output behavior.
LLM Optimization
LLM Agent Automates Hardware-Aware Model Quantization
New research introduces an LLM-based agent that automatically selects optimal quantization strategies for deploying large language models across diverse hardware platforms.
LLM fine-tuning
RVPO: Stabilizing LLM Fine-Tuning Through Variance Control
New research introduces Ratio-Variance Regularized Policy Optimization (RVPO), a method that stabilizes reinforcement learning from human feedback by controlling importance sampling variance in LLM training.
LLM Training
SIGMA: Spectral Analysis Reveals Early Warning Signs of LLM Colla
New research introduces SIGMA, a scalable spectral method using eigenvalue analysis to detect model collapse during LLM training before performance degrades catastrophically.
AI Agents
Context Engineering: The New Discipline Powering AI Agents
Beyond prompt engineering, context engineering is emerging as the critical discipline for building reliable AI agents—managing what information models see, when, and how.