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

3,367

matching records

Shown

current page

Top radar

http://arxiv.org/abs/2610.12449v1

Code links

on this page

Paper	Authors	arXiv ID	Categories	Published	Code	Radar	Summary
Bi-FORK: Generative Modeling of High-Dimensional Bifurcating Systems	Anna Zimmel, Fleur Hendriks, Markus Holzleitner	http://arxiv.org/abs/2610.12449v1	cs.LG, cs.AI	Oct 8, 2026	None	Score72	Bifurcations are ubiquitous in physical systems, from structural buckling to fluid and climat...
Rounding in Preconditioner Space: Redesigning 4-bit AdamW Optimizer-State Quantization	Hanyang Li, Shao Tang, Daniel Thomas Braithwaite	http://arxiv.org/abs/2610.12444v1	cs.LG	Oct 8, 2026	None	Score77	Quantizing AdamW's optimizer states reduces persistent storage, but quantization errors propa...
A Unified Bellman Operator for Safety-Critical Reinforcement Learning	Nishanth Arun Rao, Royina Karegoudra Jayanth, Benjamin Eysenbach	http://arxiv.org/abs/2610.12420v1	cs.LG	Oct 8, 2026	None	Score75	Reinforcement learning in safety-critical domains requires maximizing task performance while...
Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs	Zhewei Chen, Hao Zhu, Jiaojiao Jiang	http://arxiv.org/abs/2610.10520v1	cs.LG	Oct 7, 2026	None	Score74	GNN-to-MLP distillation aims to retain the predictive accuracy of a message-passing teacher w...
Your Prompt Should Do More: Effects of Retrieval Instructions in Embedding Models	Amanda Myntti, Jenna Kanerva, Veronika Laippala	http://arxiv.org/abs/2610.10508v1	cs.CL	Oct 7, 2026	None	Score83	Prompted embedding models have recently received increasing attention, particularly for retri...
A Society of Researchers: Designing Institutions for Populations of Autonomous Research Agents	Ali Asaria, Deep Gandhi, Tony Salomone	http://arxiv.org/abs/2610.10468v1	cs.MA, cs.AI	Oct 7, 2026	None	Score73	Deployments of research agents are moving to populations of thousands that share one pool of...
ALIVE: Interaction-Aligned Object Insertion for First-Frame-Guided Video Editing	Zhenghong Zhou, Zhe Lin, Jiebo Luo	http://arxiv.org/abs/2610.08779v1	cs.CV	Oct 6, 2026	None	Score74	Current video editors can insert objects but often struggle to make them participate in inter...
Post-Training Semantic Lifting for 3D Gaussian Splatting: Separating Detector, Lifting and Representation Error	Iván Verdugo Guerra, Ezequiel López Rubio, Jorge García González	http://arxiv.org/abs/2610.08756v1	cs.CV	Oct 6, 2026	None	Score80	The same Gaussian of a 3D Gaussian Splatting model is seen from many views, and these views d...
Deep Learning for Sleep Heart Rate Estimation from Accelerometers: Toward Population-Scale Cardiac Insight Without Optical Sensors	Tanbin Islam Rohan, Pranjol Sen Gupta, Tanusree Debi	http://arxiv.org/abs/2610.06823v1	cs.LG, cs.AI	Oct 5, 2026	None	Score80	Large longitudinal cohorts often contain wrist accelerometry without optical heart-rate sensi...
Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice Text-to-Speech Model	Sahil Mahendrakar	http://arxiv.org/abs/2610.06817v1	cs.SD, cs.AI	Oct 5, 2026	Detected	Score81	We distill Kokoro-82M, a widely used open text-to-speech model with 54 voices, into Paradee,...