Optimization over covariance matrices with a parameterized metric
Two-parameter Riemannian metric family unifies Euclidean, Bures-Wasserstein, and affine-invariant metrics for covariance matrix optimization.
Search the full wire by company, model, lab, or keyword. Every story we have ever aggregated.
Two-parameter Riemannian metric family unifies Euclidean, Bures-Wasserstein, and affine-invariant metrics for covariance matrix optimization.
First systematic OOD evaluation of congenital heart disease segmentation across scanner/protocol/institution/modality shifts using ImageCHD.
Human-centered semantic validation framework for network traffic classification models to detect spurious pattern exploitation beyond predictive metrics.
Hyperbolic geometry framework for genomic sequence representation using Lorentz encoding; domain-specific bioinformatics.
Directional decomposition analysis reveals parallel components in transformer representation evolution beyond residual paths, with space-dependent asymmetry in attention/MLP updates.
KnowBench introduces Effort Reduction metric for evaluating clinical AI systems based on clinician acceptance of auto-generated work products rather than reference similarity.
Anthropic CEO Dario Amodei says the time has come to slow down AI development and will give third-party evaluators like METR access to its models to help ensure its "adherence to safety practices and commitments." In a winding essay, Amodei proposed a three-step plan to "pace…
MAxBench benchmark for evaluating fine-grained steering and representation geometry recovery of multinomial concepts in language models.
Groupoid-based RL framework captures local, state-dependent symmetries to exploit modularity in realistic environments.
GPU utilization analysis reveals Nvidia Hopper's SM metrics hide inefficiency in LLM decode phase due to small-batch matrix fragmentation.
PA-CDM introduces position-aware character detection matching for handwritten math expression recognition, improving position-blind metrics.
Weight spectral density metrics predict membership inference attack vulnerability without shadow models, enabling efficient large-scale privacy auditing.
Controlled inversion test measures whether LLMs can undo news framing (lexis, agency, salience) while preserving facts, revealing generation-detection asymmetries.
Language-augmented semantic priors integrate NLP design intent into B-spline surface fitting for CAD geometry kernels.
Framework combining symbolic geometry parser with LLM solver to match LMM performance on plane geometry problems with reduced opacity.
DoRA fine-tuning approach incorporating individual text corpora to personalize small language models for knowledge QA tasks.
Mathematical generalization of Amari's Bayesian duality via convex analysis, connecting classical information geometry to modern AI.
Knowledge graph-based evaluation framework to assess contextual understanding in LLMs beyond surface-level metrics.
SQLMorph: query mutation framework for reproducible Text-to-SQL evaluation on complex enterprise schemas.
Fact-Ablated Evaluation framework audits whether LLMs faithfully use evidence or rely on parametric knowledge in fact-checking.
Robustness audit via NC-MCAR metric reveals LMs destabilize answers when source-attributed cues contradict evidence.
Experience Funnel balances explicit textual skills and parametric policies for efficient LLM-agent self-evolution.
BatchNorm artifacts confound machine unlearning evaluation: single forward pass on retain data reverses forgetting metrics.
Silent Revision: metric measuring undisclosed changes in frontier AI safety frameworks; corpus of versioned developer commitments.
Machine translation evaluation metrics accounting for terminology variation in English-French scientific translation; cross-term variation diagnostic.
Proposes role-specific predictive geometries for nonstationary multivariate graph-signal forecasting using separate long-run and short-run directional relations.
Study characterizes interaction between inference-time activation steering and weight-only quantization (INT8, NF4) on 7-9B open-weight LLMs.
Self-supervised and generative pretraining improve deep learning for medical image distortion correction in diffusion-weighted imaging.
Routing effective rank metric reveals reproducible low-high-low trajectory in MoE expert routing across test-time scaling, with implications for inference budget allocation in math/science tasks.
Activation steering in LLMs validates whether steering vectors encode coherent human-value geometry using Schwartz's moral theory.