Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Researchers proposed HERMES, a graph-based framework that predicts patient outcomes using only clinical text, building personalized knowledge graphs via Large-Language-Model-guided extraction with Contrastive Logic Model…
A three-stage ASR pipeline built on Qwen2.5-Omni-3B achieved 80-85% entity recall and 76-86% filler recall across 6,000 test utterances from speakers in India, Indonesia, and Latin America, up from 53-55% and under 5% re…
A new arXiv paper (2609.20839v1) proposes progressive error curriculum training (PECT), a curriculum learning framework that adapts a No Language Left Behind (NLLB)-based phoneme-to-text reconstruction model using synthe…
COAL-SQL, a unified post-training framework combining Coverage-Guided Augmentation (CGA) and Failure-Driven Learning (FDL), achieves 64.9% execution accuracy on the BIRD development set using only 12,600 distinct post-tr…
Researchers proposed VISPATH, a visual-intent-guided path reasoning framework for multimodal knowledge graph question answering (MM-KGQA), according to a new arXiv paper (arXiv:2609.20843v1). VISPATH recomputes hop-speci…
A new arXiv paper (2609.20844v1) reports that DLD-RL, a three-stage training method that converts Deepresearch reinforcement-learning rollout trajectories into long-context QA instances at zero annotation cost, outperfor…
A new arXiv paper (2609.20846v1) reports that fine-tuning several 4B large reasoning models with a novel GRPO reward that encourages efficient reasoning about whether a task contains enough information improves abstentio…
Researchers introduced Elastic Threshold Attention (ETA), an end-to-end trainable sparse attention architecture that predicts dynamic, contextual thresholds from query representations to speed up long-context decoding wi…
Researchers introduced Bio-MF, a latent-free one-step MeanFlow framework for EEG-conditioned fNIRS generation, detailed in arXiv paper 2609.20904v1. On Dataset 1, EEG plus synthetic fNIRS improved accuracy over EEG-only …
Researchers introduced Stiefel-AdamW, a near drop-in replacement for the AdamW optimizer that constrains one factor of linear factorization blocks to the Stiefel manifold, according to arXiv paper 2609.21039v1. The metho…
Researchers introduced EnSol, an environment-aware probabilistic graph neural network for molecular solubility prediction, which achieved Spearman correlations of 0.876 on the SolProp benchmark and 0.601 on the Leeds ben…
A new arXiv paper (arXiv:2609.21126v1) proposes a decoupled, layerwise method for structurally sparsifying the fully connected layers of pretrained neural networks, extracting shallow two-layer subnetworks, normalizing i…
A new neural network model for identifying text content file types, especially source code, is approximately four times faster than Google's Magika and 28% smaller in size, according to an arXiv paper (2609.21306v1). The…
MIRCID, a framework comparing gene expression with inferred transcription factor activity and miRNA expression, infers 414 pan-cancer hub miRNAs from 977 L1000 landmark genes with a Pearson correlation coefficient of 87.…
A new arXiv paper (2609.21309v1) surveys and formalizes strategies for using lossless compression algorithms such as gzip as machine learning methods, via Normalized Compression Distance or the Minimum Description Length…
A new arXiv paper (2609.21327v1) introduces Deep-BQRL, a model-free distributional reinforcement-learning framework that extends buffered-quantile learning to neural function approximation, learning conditional return qu…
A new arXiv paper (2609.21425v1) argues that zero-shot time-series forecasting should be governed as an evidence-access claim rather than a training-status condition, since a frozen language model prompted with serialize…
A new arXiv paper (2609.21382v1) presents a generative sequence model for business process forecasting that uses marked temporal point processes to produce full predictive distributions over the next activity and its int…
A new arXiv paper (2609.20912v1) reports that RydbergGPT, an autoregressive transformer trained on qubit projective measurement data from interacting Rydberg atom arrays, follows a power-law loss-versus-dataset-size rela…
A new arXiv paper (2609.20883v1) introduces the class of "sparse priors" and defines "Sparse Dimension" as a measure of prior sparsity over the space of all distributions, showing that distribution learning under a k-spa…