Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.
CoRe-MARL applies recurrent multi-agent reinforcement learning to cooperative supply redistribution among decentralized emergency response centers.
SK Hynix told TechCrunch the company hasn't finalized any plans or arrangements yet.
GenStream replaces dense video frames with compact semantic representations of human-centric content to reduce streaming bandwidth via generative reconstruction.
Prelabeling bounds and linear programs quantify minimum label requirements to distinguish selective prediction performance among identical-prediction classifiers.
Conditional Neural Manifolds replace fixed array manifolds with input-conditioned mappings for robust direction-of-arrival estimation under model mismatch.
Weakening neurons in GLU-based LLMs suppress residual stream directions via negative input-output cosine similarity, concentrated in late layers across models.
Geometric theory of decision-boundary structures in Markov decision processes for understanding policy complexity and oracle-query efficiency.
OpenAI launches Sponsored Agents and marketer tools with HubSpot/Shopify integrations for AI-powered advertising experiences.
The Palo Alto startup says it has landed multiple seven-figure enterprise contracts within months of launch.
PACT benchmark evaluates enterprise LLM agents' compliance under pressure from users in hiring, healthcare, and finance contexts.
Robust Monte Carlo Tree Search variant mitigates simulator-to-reality dynamics mismatch for real-world sequential decision problems.
SynAgent uses multimodal LLM agents to autonomously guide materials synthesis experiments with explicit reasoning and adaptive analysis.
MAGER uses multi-agent genetic evolution to auto-discover reasoning paths over propagation graphs for zero-shot fake news detection.
ReDIL-GNN framework for domain-incremental learning in circuit GNNs with resynthesis adaptability scoring for logic design.
Study separates recursive reasoning from statistical extrapolation in LLM in-context learning using multi-agent game experiments.
The AI boom is becoming a materials challenge. As AI pushes computing into new territory, the materials behind that infrastructure are becoming just as crucial as the algorithms running on it. Semiconductors and data centers are approaching physical limits around performance, thermal management, electrical efficiency, and reliability, creating new demands for materials that can do…
Peak-aware short-term load forecasting for distribution grid operators across UK and Swiss aggregation levels.
Label-free test-time RL optimizes only ~100K bias parameters for reasoning improvement; achieves MATH-500 gains without labeled data.
TTM-Bench standardizes benchmarking protocol for text-to-music systems with unified metrics across architectures and conditioning modes.
On-the-fly homography calibration for multi-camera tracking eliminates offline 3D site calibration via geometric refinement.
Recursive TensorSketch improves trace estimation efficiency for large implicit matrices with fewer random bits.
Physics-informed Neural ODEs reconstruct AdS black hole spacetime from boundary spectral functions via holographic duality.
NISQ-compatible quantum transformer with variational quantum blocks replaces attention layers for four-token sequence modeling.
Population genetics framework interprets AI model generations as sexual/asexual reproduction across LLMs and neural networks.
Patch-based U-Net, ResNet-50, physics-informed, and Swin-Unet surrogates predict wildfire spread from 10k+ ensemble simulations.
Fourier-based semidefinite relaxation formalizes echo chamber detection as distinct from community detection and graph partitioning.
Iterative algorithm provably identifies terminal nodes and reconstructs DAGs in structural equation models with latent confounders.
Probe-based representation steering fails to reduce WER gaps across speaker groups in Whisper, HuBERT, Wav2Vec2 encoders.
Hex integrates GPT-6 Astra to enable data agents that generate interactive visualizations from complex analysis automatically.
Over the past few days, a lot of people who stand to make a lot of money from AI all publicly agreed that it's time to make everyone slow down before we lose control - including OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Google DeepMind cofounder Demis Hassabis, Microsoft CEO Satya Nadella, and X CEO Elon Musk. When people who profit from something declare that it's dangerous and needs to be regulated, there's always reason to be skeptical. But whatever their reason, this is far from the first time AI thought leaders have sounded the alarm. Charles Darwin's theory of evolution prompte...