PulseAugur / Brief
EN
LIVE 15:05:19

Brief

last 24h
[3/3] 224 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. TOOL · arXiv cs.LG English(EN) ·

    Spectral DPPs via NEPv: A Scalable Continuous Relaxation of Determinantal MAP for Diversity-Aware Data Selection

    Researchers have developed a new method called Spectral DPPs via NEPv to address the NP-hard problem of selecting diverse, high-quality subsets from large datasets. This approach recasts the Determinantal MAP objective as a continuous optimization problem on the Stiefel manifold, leading to a Nonlinear Eigenvalue Problem with eigenvector dependency (NEPv). The proposed solver, OurMethod, offers a scalable solution that integrates with common machine learning kernels and scales near-linearly with the size of the candidate pool. AI

    Spectral DPPs via NEPv: A Scalable Continuous Relaxation of Determinantal MAP for Diversity-Aware Data Selection

    IMPACT This method could improve efficiency in data curation and subset selection for training large AI models.

  2. TOOL · arXiv cs.LG English(EN) ·

    M\=oLe-{\Lambda}: Learning the Coupled-Cluster Response State for Energies, Gradients, and Properties

    Researchers have developed MōLe-Λ, a novel machine learning model designed to predict quantum chemistry properties more efficiently. This model extends the existing MōLe framework to learn both the right-hand (T) and left-hand (Λ) amplitudes of the coupled-cluster (CC) response state. By jointly learning these amplitudes from localized molecular orbitals, MōLe-Λ can accurately predict energies, forces, dipoles, quadrupoles, polarizabilities, and electron densities, offering a significant speed advantage over traditional CCSD calculations. AI

    M\=oLe-{\Lambda}: Learning the Coupled-Cluster Response State for Energies, Gradients, and Properties
  3. TOOL · arXiv cs.LG English(EN) ·

    Excited Pfaffians: Generalized Neural Wave Functions Across Structure and State

    Researchers have developed a novel neural network architecture called Excited Pfaffians, designed to more efficiently represent multiple quantum states. This approach significantly reduces computational cost compared to traditional methods, enabling faster training and the modeling of a greater number of states. The architecture has successfully been applied to complex systems like the carbon dimer and the beryllium atom, marking a first for neural network applications in these areas. AI

    Excited Pfaffians: Generalized Neural Wave Functions Across Structure and State

    IMPACT Introduces a novel neural network architecture that significantly accelerates quantum state calculations, potentially enabling new discoveries in computational chemistry and physics.