Research radar

arXiv preprints, filtered for developer impact.

Concise summaries of new AI papers, ranked for how likely they are to change how production systems are built. arXiv collection remains rate-limited.

Papers tracked

180

1 arXiv sources

Top radar

http://arxiv.org/abs/2605.30352v1

Published 24h

by paper date

Rate limit

1/3s

arXiv policy

Paper	Authors	arXiv ID	Categories	Published	Code	Radar	Summary
On-Policy Adversarial Flow Distillation for Autoregressive Video Generation	Yang Luo, Shengju Qian, Xiaohang Tang	http://arxiv.org/abs/2605.26105v1	cs.CV	May 25, 2026	None	RDR61	Autoregressive video generators are attractive for streaming, long-horizon, and interactive a...
On the Stability of Spherical Hellinger-Kantorovich Flows and Their Implications for Differential Privacy	Aratrika Mustafi, Soumya Mukherjee	http://arxiv.org/abs/2605.23879v1	stat.ML, cs.CR	May 22, 2026	None	RDR61	Gradient-flow sampling interprets a Gibbs distribution as the minimizer of an energy function...
Tokenisation via Convex Relaxations	Jan Tempus, Philip Whittington, Craig W. Schmidt	http://arxiv.org/abs/2605.22821v1	cs.CL, cs.LG	May 21, 2026	None	RDR61	Tokenisation is an integral part of the current NLP pipeline. Current tokenisation algorithms...
Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models	Krishnakumar Balasubramanian	http://arxiv.org/abs/2605.22795v1	stat.ML, cs.AI	May 21, 2026	None	RDR61	We propose and analyze a conservative drifting method for one-step generative modeling. The m...
Uncertainty-driven 3D Gaussian Splatting Active Mapping via Anisotropic Visibility Field	Shangjie Xue, Jesse Dill, Dhruv Ahuja	http://arxiv.org/abs/2605.30342v1	cs.CV, cs.RO	May 28, 2026	None	RDR60	We present Gaussian Splatting Anisotropic Visibility Field (GAVIS), a novel framework for unc...
HarmoVid: Relightful Video Portrait Harmonization	Jun Myeong Choi, Jae Shin Yoon, Luchao Qi	http://arxiv.org/abs/2605.28811v1	cs.CV	May 27, 2026	None	RDR60	We present a method for harmonizing the lighting of a foreground video to match a target back...
TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstruction	Weijie Wang, Zimu Li, Jinchuan Shi	http://arxiv.org/abs/2605.26115v1	cs.CV	May 25, 2026	None	RDR60	Sparse-view 3D reconstruction is increasingly addressed with feed-forward splatting networks...
Vision Transformers Need Better Token Interaction	Linxiang Su	http://arxiv.org/abs/2605.23868v1	cs.CV	May 22, 2026	None	RDR60	Vision Transformers (ViTs) can learn strong image-level representations while their patch rep...
GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction	Katharina Schmid, Nicolas von Lützow, Jozef Hladký	http://arxiv.org/abs/2605.23888v1	cs.CV	May 22, 2026	None	RDR58	We introduce a new approach to high-fidelity 3D scene reconstruction from multi-view RGB imag...
LaMo: Self-Supervised Latent Motion Priors for Physical Realism in Video Generation	Bo Jiang, Depu Meng, Yihan Hu	http://arxiv.org/abs/2605.23878v1	cs.CV	May 22, 2026	None	RDR58	Modern video generators produce visually compelling clips but still struggle with physical an...