🏢 Johns Hopkins University

DINeMo: Learns 3D models with no 3D annotations, leveraging pseudo-correspondence from visual foundation models for enhanced pose estimation.

AgentRxiv enables collaborative autonomous research via LLM agent preprint sharing, boosting performance and discovery.

R2-T2: Boost multimodal MoE performance by re-routing experts in test-time, no retraining needed!

Test-time scaling + confidence = better QA!

Smaller image patches improve vision transformer performance, defying conventional wisdom and revealing a new scaling law for enhanced visual understanding.

GenEx generates explorable 3D worlds from a single image, enabling embodied AI agents to explore and learn.

Generative World Explorer (Genex) enables agents to imaginatively explore environments, updating beliefs with generated observations for better decision-making.