Wrapped "Urban Networks in the Age of AI" at #NetSci2026. Career shifts predate ChatGPT (Frank), CHIPS fabs cluster but workers live elsewhere (Farzaneh), data centers reshape cities (Liang), frontier AI = talent circulation (Chang).
More on full program: https://t.co/G4eRcAzG2G
We are excited to launch AutoSOTA Live @ CVPR 2026: an attempt to reproduce and improve the latest CVPR papers at scale during the conference.
Autonomous research systems have recently shown great potential to accelerate AI research, such as AutoResearch by @karpathy and AI Scientist by @SakanaAILabs . Inspired by these breakthroughs, AutoSOTA aims to scale automated AI research to the level of top AI conferences: while CVPR presents new ideas, AutoSOTA stress-tests, reproduces, and pushes them further in real time. AutoSOTA does not seek to diminish the original scientific contributions of human researchers; rather, it aims to serve as an innovation accelerator, exploring how far these ideas can be advanced within a very short time period.
As AI research continues to scale, reproducing, validating, and extending published methods has become an increasingly important part of the research cycle. AutoSOTA closes this loop through a tightly coordinated multi-agent architecture designed to mirror the workflow of human research teams. Over the coming days, AutoSOTA will progressively produce detailed traces of replication and enhancement for the latest CVPR papers, offering full transparency into how automated agents diagnose code, reflect on model architectures, and improve performance beyond the original SOTA.
π Paper: [https://t.co/lLaFFja3pG]
π Live Updates: [https://t.co/w3XcHD7hjI]
#AutoSOTA_LiveCVPR #CVPR2026 #AutoSOTA #AIScientist #AutoResearch #MachineLearning #AcademicX #ResearchAutomation
We especially encourage work grounded in real urban applications, including transportation, navigation and routing, recommender systems, on-demand labor platforms, emergency response and urban resilience, public health, housing and neighborhood change, energy and sustainability.
The city each person lives in is distinct...shaped by places they can and cannot go, experiences they can and cannot have, people with whom they can and cannot connect, and opportunities--for prosperity, perspective, safety and more--they can and cannot pursue. Inequality and segregation begin with differential *access* to the city. They reflect objectively subjective phenomena with vast and enduring material consequences. In a paper out today in Nature Human Behaviour (https://t.co/8i4ayoKTIP), my amazing colleagues (+ moi) demonstrate how to measure and manage these subtle but profoundly impactful realities with new digital data on mobility across the complex graph of urban places and people. Check it out! Led by the amazing @xu_fengli, with Qi Wang, @estebanmoro, @lin_chen_tweet, @_arianna_sm, @martikagv, @mtizzoni@chaomingsong, Carlo Ratti @senseablecity, @bettencourtluis, and Yong Li.
I am looking for postdoc opportunities starting in 2025. If you know of any positions related to social computing/urban science/urban mobility/LLM agents, I would greatly appreciate your references or recommendations!
Found myself in the third picπWelcome to check out our paper, "Large Language Model-driven Meta-structure Discovery in Heterogeneous Information Network"!
Paper: https://t.co/5UY3wo49EY
Code: https://t.co/pLtx71oQUm
2-min promotional video: https://t.co/KPRKVyeUvm
#IJCAI2024 AI for Social Good #Talk VulnerabilityMap: An Open Framework for Mapping Vulnerability among Urban Disadvantaged Populations in the United States by Lin Chen, Yong Li, Pan Hui
@feedkoko@daria_kravets@IC2S2 Thanks for all your devotion! I also see my results in OpenReview but have not received the notification email yet. Could you also check my case?
Excited to share that I've successfully pulled off a double winβgot my submission accepted to @IC2S2, and got myself accepted to @SICSS_Penn! Can't wait to engage in both sparkling events!