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/2610.13895v1

Code links

on this page

Paper	Authors	arXiv ID	Categories	Published	Code	Radar	Summary
Task Context Shapes Component Roles in Language Models	Sovesh Mohapatra, Dani S Bassett	http://arxiv.org/abs/2610.13895v1	cs.CL, cs.AI	Oct 9, 2026	None	Score63	Large language models reuse the same attention heads across tasks, but identifying these shar...
Rubric-CEPR: Self-Evolving Image Editing via Reward-Verified Self-Distillation	Ritesh Thawkar, Shubham Patle, Shravan Venkatraman	http://arxiv.org/abs/2610.12469v1	cs.CV	Oct 8, 2026	None	Score75	Instruction-guided image editors have become highly capable, yet improving them further still...
Ecology of AI Agents: Collaboration Creates a Population Threshold for Takeoff	Erin Crawley, Hidenori Tanaka	http://arxiv.org/abs/2610.12436v1	cs.AI, cond-mat.dis-nn	Oct 8, 2026	None	Score62	AI agents can now conduct real-world cyberattacks, scale up capabilities with the number of a...
Never Look Back: Understanding Persistence in 3D Object Memory from Egocentric Videos	Shravan Chaudhari, William Paul, Suchi Saria	http://arxiv.org/abs/2610.10538v1	cs.CV, cs.AI	Oct 7, 2026	None	Score74	As we move through the world and carry out everyday tasks, we encounter objects that may beco...
Video-Conditioned Generative Joint 2D-3D Hand Motion Recovery	Chen Xu, Yunqi Li, Binbin Huang	http://arxiv.org/abs/2610.10512v1	cs.CV	Oct 7, 2026	None	Score69	Recovering faithful 3D hand motion from video remains challenging due to frequent occlusions...
One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline	Shih-Chen Tseng, Chih-Hsuan Chen, Ryan Yang	http://arxiv.org/abs/2610.06852v1	cs.CV, cs.AI	Oct 5, 2026	None	Score81	Pipeline figures in ML papers must be repurposed across many canvases, including paper column...
MemPilot: Orchestrating On-Demand Multimodal Memory Curation for LLM Agents	Haozhen Zhang, Haodong Yue, Quanyu Long	http://arxiv.org/abs/2610.06830v1	cs.CL, cs.AI	Oct 5, 2026	None	Score75	Memory has become integral to the LLM agent ecosystem, supporting information retention and r...
Less Context, Better Geometry: Masked Geometric Encoder for Robust 3D Foundation Models	Zhimin Shao, Xijun Liu, Zhaoliang Zhang	http://arxiv.org/abs/2610.06813v1	cs.CV	Oct 5, 2026	None	Score77	Recent progress in 3D foundation models has enabled rapid 3D reconstruction and camera calibr...
Sharpen Without Search: On-Policy Distillation of Sequence-Level Power Distribution	Erfan Baghaei Potraghloo, Seyedarmin Azizi, Arya Fayyazi	http://arxiv.org/abs/2610.06804v1	cs.LG, cs.AI	Oct 5, 2026	Detected	Score85	A language model can give a correct answer more probability than any single incorrect answer...
Usefulness of Quantile-Aware Diffusion Modeling for Highly Imbalanced Tabular Data	Abu Talha, Peng Liu, Souradyuti Paul	http://arxiv.org/abs/2610.05825v1	cs.LG, stat.AP	Oct 5, 2026	None	Score80	Classification problem in the context of highly imbalanced data is a major challenge in many...