The Archive

Moral Entropy framework decomposes annotator disagreement into aleatoric and epistemic uncertainty for computational ethics tasks.

Spectrally-aligned latent flow matching improves time series generation by preserving dynamical properties in synthetic data.

MORM introduces multiplicatively optimistic regret matching for finite general-sum games with sublinear external regret bounds.

Manus, which earlier this year had to break off a merger with Meta, is in discussions to raise $500M at a $4B valuation.

As the tech sector sounds alarms about AI's potential to destroy humanity, entertainment labor groups are urging the public to stay focused on what's already happening. The Verge reached out to Disney, Netflix, Amazon, Lionsgate, and other studios who have started using AI, as well film startups focused on bringing generative AI into the mainstream to ask for their reaction to the recent warnings surrounding the technology. None have responded to our request for comment. The Screen Actors Guild - American Federation of Television and Radio Artists (SAG-AFTRA) and Writers Guild of America East...

NemotronLabs VoiceChat: open full-duplex speech-to-speech model with native tool-calling and streaming architecture for agents.

ECG biometrics using Siamese ResNet evaluated across exercise stress and temporal variation for authentication and pretext learning.

Analysis of tau-leaping schedule optimization in masked discrete diffusion models with bounds on factorization error.

RACER enables role-aligned competence estimation for human-AI routing with instance-level expert specialization modeling.

Shared latent-space framework connects simulator calibration and RL control for urban traffic optimization and real-time management.

Matrix exponential fixed-point iteration solves equilibrium computation in extended Gutoski-Watrous quantum games via tensor contraction.

Anthropic partners with Accenture to embed model evaluation capabilities into enterprise workflows.

Benchmark for robot failure diagnosis: sensor evidence auditing determines when robots should ask humans vs. act autonomously.

Hard-label model extraction attack on ReLU MLPs via efficient sign recovery, extending Eurocrypt 2025 Carlini et al. work.

AutoViewMem framework organizes LLM agent long-term memory into orthogonal semantic views to reduce retrieval noise.

Cross-expert neural framework for lithium-ion battery RUL prediction and capacity estimation from partial charging data.

Identifiability proof for real analytic nonlinear ICA with Laplace-like source distributions via discontinuity analysis.

RAVEL: reinforcement learning framework for query selection in interactive retrieval under partial evidence.

Nonparametric tangent vector field regression on Riemannian manifolds via parallel transport and kernel methods.

Benchmark comparing HyFyDy and MuJoCo motion-imitation RL pipelines on musculoskeletal fidelity vs kinematic accuracy.

Clinical world models require intervention bundle granularity: MIMIC-IV analysis shows treatment components co-occur atomically.

Nvidia's Nader Khalil and Sydney Sykes discuss one of the decisions shaping next-gen startups on the Builders Stage at TechCrunch Disrupt 2026.

ExpBoN: soft best-of-n LLM alignment via exponential-noise mechanism with exponentially fast KL-regularized convergence.

Muse is now available on the Mac, where it can work with your files and apps to take action on your behalf.

Iterative framework using generator and extractor LLMs to synthesize interpretable categorical features for tabular prediction.

Novel membership inference detector using entropy-corrected loss boundaries to distinguish pretraining data exposure from learned generalization in LLMs.

First-order optimization algorithm for Hölder-smooth convex objectives with theoretical acceleration bounds in high dimensions.

Benchmark framework evaluating explanation quality alongside accuracy for vision-language models in forensic face recognition tasks.

Geometric Mean Pooling operator combining sign and magnitude information for hierarchical neural network feature composition without learnable parameters.

Adaptive spatio-temporal neural field using learned tessellations to represent variable-complexity climate patterns across geographic regions.