Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
An AI 'superforecaster' designed to predict market movements, including Federal Reserve decisions, has not yet outperformed human forecasters, according to a Financial Times report. The AI's performance highlights curren…
A new paper provides an accessible overview of sheaf theory applications in deep learning, data science, and computer science, bridging classical mathematics with modern implementations in signal processing and deep lear…
Researchers introduced Lipschitz Scaling Training (LiST), a method that adjusts the Lipschitz constant of neural networks to achieve out-of-the-box calibration while balancing accuracy and robustness. LiST outperforms ba…
Researchers proposed a graph-regularized deep learning framework for EEG-based emotion recognition that models psychological interdependencies between emotion classes, achieving up to +5.42% accuracy improvement and a 39…
Researchers propose a multi-cluster boundary learning method for out-of-scope intent detection using MiniLM embedding, achieving state-of-the-art performance on CLINC150, StackOverflow, and Banking77 datasets. The method…
Researchers introduced AgentNAS, a method that uses a large language model to generate a seed architecture and decompose it into a slotted architecture for conventional neural architecture search, achieving state-of-the-…
Researchers introduced XAlpha, a memory-driven AI system that automates the end-to-end process of discovering trading signals in financial markets. The system integrates external knowledge and feedback loops to generate,…
Researchers introduced JAM, a theory-agnostic framework for personality recognition that uses an LLM-as-a-Judge mechanism to identify ambiguous samples and guide adaptive metric learning, improving cross-framework genera…
Researchers evaluated state-of-the-art tabular models as encoders in image-tabular multimodal learning for the first time, finding that In-Context Learning models, which require labels to process instances, pose a challe…
Researchers modeled seven psychological disorders in reinforcement learning agents by manipulating cognitive appraisal signals, revealing a two-dimensional affective space where disorders self-organize and interact nonad…
Researchers introduced FedTR, a federated learning framework incorporating transfer learning for industrial visual inspection, achieving 95.5% word-level accuracy on homogeneous data and 94.2% on heterogeneous data for l…
Researchers propose LOGOS, a transformer-based model that uses textual prompts to guide oriented object detection in aerial and satellite imagery. The approach outperforms existing methods on the DOTA dataset, particular…
Researchers at MIT introduced the Temporal Ratio (TR), an attention-based metric that predicts how well video-action models generalize to new tasks after finetuning on robotic data. They found that TR measures a model's …
Researchers introduced DreamCharacter-1, a lightweight post-adaptation framework that enhances pretrained 3D foundation models for high-fidelity, production-ready 3D character generation. The framework includes geometry …
Researchers identified a failure mode in reinforcement learning for large language models called Positive-Credit Contamination, where low-probability tail tokens receive undeserved positive credit. They propose TACO, a m…
Researchers propose a value-of-information router for block-sparse attention that dynamically doubles the kept key blocks when the top-k selection is uncertain, improving recall on LongBench-v2 from 0.47 to 0.75 and pres…
Researchers introduced Omni-Sleep, a foundation model for sleep analysis that leverages the physiological hierarchy of the central and autonomic nervous systems to learn structured representations from polysomnography da…
Researchers introduced ReCoLoRA, a spectrum-aware recursive consolidation method for continual fine-tuning of large language models, which prevents task interference by re-decomposing the effective weight before each new…
Researchers propose SAGA, a training-free stable acceleration guidance method for autoregressive video generation that reduces temporal errors like flickering and motion jitter by using spectral guidance and noise initia…
Researchers at an undisclosed institution applied Group Relative Policy Optimization (GRPO) to adapt an LLM-based automatic speech recognition (ASR) system to synthetic speech, achieving a 40% relative word error rate (W…