Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
AdvSynGNN proposes a Graph Neural Network architecture designed for robust node-level representation learning in the presence of structural noise and non-homophilous graph topologies.
Research explores the frequentist consistency of Prior-Data Fitted Networks (PFNs) for causal inference, addressing uncertainty quantification.
Research presents a framework to estimate accuracy and communication effort for federated perception models pre-deployment, using intrinsic data properties.
Research explores using language model's internal layer-wise trajectories, beyond maximum softmax probability, to improve uncertainty quantification.
Research proposes Boundary Embedding Shaping (BES) with adaptive contrastive learning to mitigate graph structural entanglement in GNNs, improving node classification.
Researchers propose Safety-Regulated Adaptive Transfer Reinforcement Learning (SRATRL), a teacher-student framework for safe, efficient domain adaptation.
Research explores scaling laws for contrastive representation learning using a sketched linear model, extending existing work on sketched linear regression.
Research introduces Experience Augmented Policy Optimization (EAPO) to improve LLM reasoning via Reinforcement Learning with Verifiable Rewards (RLVR).
Research introduces 'in-span learning' for reduced-order models (ROMs), allowing them to self-adapt and maintain accuracy when dynamics drift.
Research provides uniform confidence bands for Kernel Ridge Regression (KRR), improving inferential theory for nonstandard data applications.
New algorithm for linear quadratic systems achieves anytime regret of order $\mathcal{O}(\sqrt{t})$ with computational efficiency and explicit system dimension dependence.
Research introduces confidence intervals for the ℓ2 Expected Calibration Error (ECE), enhancing statistical evaluation of model calibration.
Research reviews Multi-Agent Reinforcement Learning (MARL) for adaptability, highlighting real-world violations of design assumptions.
Schrödinger Bridge Mamba (SBM) is a novel model integrating the Schrödinger Bridge training paradigm and Mamba architecture for one-step speech enhancement.
Research paper proposes "Model Gateway," a specialized MLOps platform for drug discovery, addressing multi-model composition and scientific governance.
Research proposes statistical early stopping methods for LLMs to reduce unnecessary reasoning steps, especially under uncertainty.
New research proposes a projection-augmented graph approach for Approximate Nearest Neighbor Search (ANNS) to meet modern AI workload demands.
Research explores auditable fine-tuning and inference for proprietary AI models delegated to cloud providers, addressing trust gaps with cryptographic methods.
Research introduces IConE, a self-supervised learning method preventing representation collapse without reliance on large batch sizes or batch interaction.
Research explores temporal fair division in multi-agent systems sharing limited resources, moving from precise alternation metrics to scalable coordination proxies.
Researchers adapted Quantum Leaky Integrate-and-Fire spiking neural networks (QLIF) for multivariate time-series forecasting, specifically weather.
Research proposes Q-PhotoNAS, a hybrid quantum neural architecture search for photonic devices, optimizing quantum machine learning under hardware constraints.
Research explores low-rank implicit regularization stability in deep matrix factorization, clarifying conditions for gradient descent to favor low-rank structures.
Fara-1.5 is a research paper introducing FaraGen1.5, a scalable data pipeline for training computer use agents with environments, solvers, and verifiers.
Research details methods for generating test inputs for tensor programs to find kernel bugs, evaluating seven strategies across 26 operations.
Research explores if randomness is essential for adaptive data analysis to prevent overfitting and false discovery when reusing datasets.
A research paper details large-scale development and validation of a deep learning system, Inspectra CXR, for thoracic disease detection on chest radiographs in Thailand.
Research proposes a Conversational Risk Accumulation (CRA) framework with stateful guardrails to address safety failures in multi-turn LLM systems.
Research finds Supervised Fine-Tuning (SFT) in LLMs reduces behavioral diversity in sequential decision-making tasks, potentially narrowing reasoning.
Research explores using hypernetworks for train-time knowledge injection into LLMs by generating LoRA adapters for large factual corpora.
© 2026 OneBench: AI Insights. All rights reserved.
Evidence before opinion