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
Unlocking the Working Memory of Large Language Models for Latent Reasoning	Lukas Aichberger, Sepp Hochreiter	http://arxiv.org/abs/2605.30343v1	cs.CL, cs.AI	May 28, 2026	None	RDR76	To improve the reasoning capabilities of large language models, test-time compute is typicall...
CaMBRAIN: Real-time, Continuous EEG Inference with Causal State Space Models	Abhilash Durgam, Nyle Siddiqui, Jeffrey A. Chan-Santiago	http://arxiv.org/abs/2605.28792v1	cs.AI, cs.HC	May 27, 2026	None	RDR76	Electroencephalography (EEG) is a critical, non-invasive method to monitor electrical brain a...
Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use Agents	Suji Kim, Kangsan Kim, Sung Ju Hwang	http://arxiv.org/abs/2605.28775v1	cs.LG, cs.AI	May 27, 2026	None	RDR76	Computer-use agents (CUAs) have recently made substantial progress, but deploying a separate...
GENESIS: Harnessing AI Agents for Autonomous 6G RAN Synthesis, Research, and Testing	Tamerlan Aghayev, Maxime Elkael, Michele Polese	http://arxiv.org/abs/2605.27360v1	cs.NI, cs.AI	May 26, 2026	None	RDR76	Cellular research and development (R&D) is throttled by six structural processes that each co...
Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders	Yi Jing, Zao Dai, Jinwu Hu	http://arxiv.org/abs/2605.27354v1	cs.LG, cs.AI	May 26, 2026	None	RDR76	Model internals encode rich information about how a large language model (LLM) processes its...
Feedforward 3D Editing Learns from Semantic-Part Transformation	Jiawei Weng, Saining Zhang, Zhenxin Diao	http://arxiv.org/abs/2605.27351v1	cs.CV	May 26, 2026	None	RDR76	3D editing is a fundamental capability for scalable 3D content creation. While image editing...
CHRONOS: Temporally-Aware Multi-Agent Coordination for Evolving Data Marketplaces	Joydeep Chandra	http://arxiv.org/abs/2605.23887v1	cs.DB, cs.AI	May 22, 2026	None	RDR76	Temporal knowledge-graph data marketplaces face three coupled failures in static designs: sta...
Integrable Elasticity via Neural Demand Potentials	Carlos Heredia, Daniel Roncel	http://arxiv.org/abs/2605.22820v1	cs.LG	May 21, 2026	None	RDR76	We propose the Integrable Context-Dependent Demand Network (ICDN), a demand-first neural mode...
Locally Coherent, Globally Incoherent: Bounding Compositional Incoherence in Multi-Component LLM Agents	Anany Kotawala	http://arxiv.org/abs/2605.30335v1	cs.AI, cs.CL	May 28, 2026	None	RDR75	Multi-component LLM agents assemble probabilistic claims from components that each see only p...
Gram: Assessing sabotage propensities via automated alignment auditing	David Lindner, Victoria Krakovna, Sebastian Farquhar	http://arxiv.org/abs/2605.30322v1	cs.LG, cs.AI	May 28, 2026	None	RDR75	We introduce Gram, an automated alignment auditing framework to assess the propensity of AI a...