AI News

June 05, 2026

OpenAI & DeepMind

AI Research (general)

(Mis)generalization of Helpful-only Fine-tuning

arXiv cs.LG (Machine Learning) · Jun 05

(Mis)generalization of Helpful-only Fine-tuning

100-LongBench: Are de facto Long-Context Benchmarks Literally Evaluating Long-Context Ability?

arXiv cs.LG (Machine Learning) · Jun 05

100-LongBench: Are de facto Long-Context Benchmarks Literally Evaluating Long-Context Ability?

2-Step Agent: A Framework for the Interaction of a Decision Maker with AI Decision Support

arXiv cs.AI · Jun 05

2-Step Agent: A Framework for the Interaction of a Decision Maker with AI Decision Support

\textsc{Lethe}: Principled Dual-Stream Update for Persistent Knowledge Erasure in Federated Unlearning

arXiv cs.LG (Machine Learning) · Jun 05

\textsc{Lethe}: Principled Dual-Stream Update for Persistent Knowledge Erasure in Federated Unlearning

A 3D Isovist World Model -- Revealing a City's Unseen Geometry and Its Emergent Cross-City Signature

arXiv cs.LG (Machine Learning) · Jun 05

A 3D Isovist World Model -- Revealing a City's Unseen Geometry and Its Emergent Cross-City Signature

A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects

arXiv cs.AI · Jun 05

A Cartography of Open Collaboration in Open Source AI: Mapping Practices, Motivations, and Governance in 14 Open Large Language Model Projects

A Finite Certificate for the Positive $n=9$ Vasc Inequality

arXiv cs.AI · Jun 05

A Finite Certificate for the Positive $n=9$ Vasc Inequality

A Framework for Measuring Appropriate Reliance on Set-Valued AI Advice

arXiv cs.AI · Jun 05

A Framework for Measuring Appropriate Reliance on Set-Valued AI Advice

A General Framework for Dynamic Consistent Submodular Maximization

arXiv cs.LG (Machine Learning) · Jun 05

A General Framework for Dynamic Consistent Submodular Maximization

A Geometric Characterization of the Stationary Plateau for Two-Layer Neural Networks

arXiv cs.LG (Machine Learning) · Jun 05

A Geometric Characterization of the Stationary Plateau for Two-Layer Neural Networks

AI in Biotech & High-Impact Journals

Hip joint image quality screening based on the Diffusion Mamba model

Nature Machine Learning · Jun 05

Hip joint image quality screening based on the Diffusion Mamba model

A universal foundation model for grounded biomedical image interpretation

Nature Machine Learning · Jun 04

A universal foundation model for grounded biomedical image interpretation

Durable Gains and Persistent Questions From the ASTRUM-005 Trial

JAMA Oncology (Online First) · Jun 04

Durable Gains and Persistent Questions From the ASTRUM-005 Trial

Explicit dynamic cross-strand interactions for DNA sequence language modelling

Nature Machine Learning · Jun 04

Explicit dynamic cross-strand interactions for DNA sequence language modelling

First-Line Serplulimab in Extensive-Stage Small Cell Lung Cancer

JAMA Oncology (Online First) · Jun 04

First-Line Serplulimab in Extensive-Stage Small Cell Lung Cancer

Fluids for Immune Checkpoint Inhibitor Diabetic Ketoacidosis

JAMA Oncology (Online First) · Jun 04

Fluids for Immune Checkpoint Inhibitor Diabetic Ketoacidosis

Fluids for Immune Checkpoint Inhibitor Diabetic Ketoacidosis—Reply

JAMA Oncology (Online First) · Jun 04

Fluids for Immune Checkpoint Inhibitor Diabetic Ketoacidosis—Reply

Georgia O’Keeffe, Hannah Wilke, and Choosing One’s Own Narrative

JAMA Oncology (Online First) · Jun 04

Georgia O’Keeffe, Hannah Wilke, and Choosing One’s Own Narrative

Neural Response to Familiar Names Predicts Outcome of Comatose ICU Patients: A Prospective Observational Cohort Study

Nature Machine Learning · Jun 04

Neural Response to Familiar Names Predicts Outcome of Comatose ICU Patients: A Prospective Observational Cohort Study

PD-(L)1 Inhibitor Monotherapy vs Chemoimmunotherapy for Advanced NSCLC With High PD-L1 Expression

JAMA Oncology (Online First) · Jun 04

PD-(L)1 Inhibitor Monotherapy vs Chemoimmunotherapy for Advanced NSCLC With High PD-L1 Expression