Each of these enormous collections of spacecraft, dubbed CTC-1 and CTC-2, would contain 96,714 satellites spread over 3,660 different orbits. If completed, China's new mega-constellation would dwarf even SpaceX's bold ambition to put 49,000 Starlink satellites in orbit. Together, CTC-1 and CTC-2 would be the largest assembly of satellites ever put in orbit, and would effectively lock competitors out of a region of low-Earth orbit. With Chinese authorities remaining quiet about the satellites' intended use, experts have raised concerns that the constellation may pose a security or defence threat. As reported by China in Space, the Nanjing University of Aeronautics claims that the satellites will focus on: 'Low-altitude electromagnetic space security, integrated security defence systems, electromagnetic space security assessment of airspace, and low-altitude airspace safety supervision services.'

Mayor blames ICE for'creating chaos' in Minneapolis and tells protesters to go home after illegal migrant is shot Iran confirms protest hero will NOT be executed as Trump reveals Tehran told him'the killing has stopped' after he threatened to take military action'Heads will roll!' US Marshals under fire for sending'Fugitive Task Force' to raid Timothy Busfield's mountain home 60 minutes after he turned himself in on child sex charges My Heated Rivalry fantasy became a reality... and my secret hookup will soon be leading the country Banks seize 367,000 homes as housing pain spreads across US... and it is about to get much worse Shocking truth about Minneapolis woman dragged from car by ICE while screaming that she was on her way to a doctor's appointment Inside Tiger Woods' 50th birthday bash as he cozies up to girlfriend Vanessa Trump to watch Bon Jovi Somali'fraudsters' force out YouTuber who exposed massive scandal amid fraud investigation in Minnesota Shameless Gwyneth Paltrow, 53, pushes'miracle' $150 Goop serum... but plastic surgeon suggests she's neglected to mention a far more invasive secret behind her taut face and'TILTED' eyes Danes say Trump is still set on'conquering' Greenland after'frank' talks as Sweden accuses him of exaggerating threats posed by Russia and China to seize the island The reclusive life of Michael Jackson's ex-wife Debbie Rowe, 67, as more details emerge over who really fathered the singer's children Palm Beach elites break out in civil war over $200m'greed project'... as Don Jr's fiancée furiously intervenes Suspended Ford worker who heckled Trump receives $600k as woke liberals call him'working class hero' White House responds to Joe Rogan's warning that ICE behavior in Minneapolis resembles Hitler's Gestapo Billionaire sports tycoon becomes America's biggest landowner with latest massive purchase Elon Musk's X stops its AI tool Grok from undressing pictures of real people after huge backlash over sexualised deepfakes MAGA goes wild for new scene in Landman with Glen Powell's glamorous love Michelle Randolph My night at America's'scariest' McDonald's that is so dangerous it does not even have a DOOR, with frightened locals renaming it'McStabby's' The REAL reason you're still single: Study reveals the biggest contributing factors - including being intelligent Whether you're happily on your own or desperate for love, you might wonder why you haven't yet met your perfect match. Now, a study has revealed the main factors contributing to remaining single - and it's bad news for those who went to university. A team from the University of Zurich recruited more than 17,000 people from the UK and Germany for their study. The participants were aged 16 at the start of the study and had no prior relationship experience. They were surveyed annually up until the age of 29 with questions aimed at capturing their characteristics, attributes and sociodemographic factors.

Mayor blames ICE for'creating chaos' in Minneapolis and tells protesters to go home after illegal migrant is shot Iran confirms protest hero will NOT be executed as Trump reveals Tehran told him'the killing has stopped' after he threatened to take military action'Heads will roll!' US Marshals under fire for sending'Fugitive Task Force' to raid Timothy Busfield's mountain home 60 minutes after he turned himself in on child sex charges My Heated Rivalry fantasy became a reality... and my secret hookup will soon be leading the country Banks seize 367,000 homes as housing pain spreads across US... and it is about to get much worse Shocking truth about Minneapolis woman dragged from car by ICE while screaming that she was on her way to a doctor's appointment Inside Tiger Woods' 50th birthday bash as he cozies up to girlfriend Vanessa Trump to watch Bon Jovi Somali'fraudsters' force out YouTuber who exposed massive scandal amid fraud investigation in Minnesota Shameless Gwyneth ...

Somber-faced Timothy Busfield appears in court for child sex abuse case as he's denied bail White House responds to Joe Rogan's warning that ICE behavior in Minneapolis resembles Hitler's Gestapo Shameless Gwyneth Paltrow, 53, pushes'miracle' $150 Goop serum... but plastic surgeon suggests she's neglected to mention a far more invasive secret behind her taut face and'TILTED' eyes Shocking truth about Minneapolis woman dragged from car by ICE while screaming that she was on her way to a doctor's appointment German troops'to touch down in Greenland in a matter of hours' as Danish leader says country is still stuck in a'fundamental disagreement' with the US over the island after'frank' meeting Benny Blanco fans left swooning after photo shows him with straight hair: 'Holy smokes' My night at America's'scariest' McDonald's that is so dangerous it does not even have a DOOR, with frightened locals renaming it'McStabby's' Progressive Portland Rep. squirms when asked about inflammatory statement she made after shooting of suspected Tren de Aragua gangsters Euphoria fans shocked over Sydney Sweeney's racy OnlyFans move: 'She's been oversexualized' Kiefer Sutherland told rideshare driver to pull over or'I'll kill you' before alleged assault in Hollywood MAGA goes wild for new scene in Landman with Glen Powell's glamorous love Michelle Randolph Six-year-old girl left all alone after ICE takes her dad away while he picked up dinner delivery: 'Where's papi?' Chicago's ultra-woke teachers' union makes glaring spelling error on flyer calling for'ultra wealthy to fund our schools' LeBron James distances himself from Rich Paul after agent pushed for Lakers to trade his client's teammates I was silenced for exposing Covid vaccine injuries in 2021... now the truth has finally come out A researcher who says she discovered that Covid vaccines could seriously injure the heart claims she was silenced during the pandemic, only to be vindicated more than four years later. Dr Jessica Rose, a Canadian researcher and expert in immunology from Memorial University of Newfoundland, said her 2021 study exposing a connection between Covid vaccines and myocarditis was mysteriously withdrawn just three weeks after it was published by the journal Current Problems in Cardiology without explanation. Myocarditis is a dangerous inflammation of the heart that can cause chest pain, shortness of breath, fatigue, irregular heartbeat, and swelling in the legs. In severe cases, it can lead to heart failure, blood clots, stroke, or sudden death. Using information from a government-run database to track vaccine side effects, Rose found a significant increase in heart damage weeks after people received the Covid vaccine.

As large language model (LLM)-based agents become increasingly integrated into daily digital interactions, their ability to reason across long interaction histories becomes crucial for providing personalized and contextually aware assistance. However, the performance of these agents in long context scenarios, particularly for action-taking WebAgents operating in realistic web environments, remains largely unexplored. This paper introduces a benchmark for evaluating long context reasoning capabilities of WebAgents through sequentially dependent subtasks that require retrieval and application of information from extended interaction histories. We develop a novel evaluation framework that simulates multi-session user interactions by injecting irrelevant task trajectories between dependent subtasks, creating contexts ranging from 25,000 to 150,000 tokens. Through extensive evaluation of four popular models, Claude-3.7, GPT-4.1, Llama 4, and o4-mini, we observe a dramatic performance degradation as context length increases, with success rates dropping from 40-50\% in baseline conditions to less than 10\% in long context scenarios. Our detailed error analysis reveals that agents primarily fail due to getting stuck in loops and losing track of original task objectives. We further propose an implicit RAG approach that provides modest improvements by generating task-relevant summaries, though fundamental limitations in long context reasoning persist. These findings highlight critical challenges for deploying WebAgents in realistic, long-term user interaction scenarios and provide insights for developing more robust agent architectures capable of maintaining coherent task execution across extended contexts.

Within Human Activity Recognition (HAR), there is an insurmountable gap between the range of activities performed in life and those that can be captured in an annotated sensor dataset used in training. Failure to properly handle unseen activities seriously undermines any HAR classifier's reliability. Additionally within HAR, not all classes are equally dissimilar, some significantly overlap or encompass other sub-activities. Based on these observations, we arrange activity classes into a structured hierarchy. From there, we propose Hi-OSCAR: a Hierarchical Open-set Classifier for Activity Recognition, that can identify known activities at state-of-the-art accuracy while simultaneously rejecting unknown activities. This not only enables open-set classification, but also allows for unknown classes to be localized to the nearest internal node, providing insight beyond a binary "known/unknown" classification. To facilitate this and future open-set HAR research, we collected a new dataset: NFI_FARED. NFI_FARED contains data from multiple subjects performing nineteen activities from a range of contexts, including daily living, commuting, and rapid movements, which is fully public and available for download.

To achieve successful assistance with long-horizon web-based tasks, AI agents must be able to sequentially follow real-world user instructions over a long period. Unlike existing web-based agent benchmarks, sequential instruction following in the real world poses significant challenges beyond performing a single, clearly defined task. For instance, real-world human instructions can be ambiguous, require different levels of AI assistance, and may evolve over time, reflecting changes in the user's mental state. To address this gap, we introduce RealWebAssist, a novel benchmark designed to evaluate sequential instruction-following in realistic scenarios involving long-horizon interactions with the web, visual GUI grounding, and understanding ambiguous real-world user instructions. RealWebAssist includes a dataset of sequential instructions collected from real-world human users. Each user instructs a web-based assistant to perform a series of tasks on multiple websites. A successful agent must reason about the true intent behind each instruction, keep track of the mental state of the user, understand user-specific routines, and ground the intended tasks to actions on the correct GUI elements. Our experimental results show that state-of-the-art models struggle to understand and ground user instructions, posing critical challenges in following real-world user instructions for long-horizon web assistance.

We develop pre-trained estimators for structural econometric models. The estimator uses a neural net to recognize the structural model's parameter from data patterns. Once trained, the estimator can be shared and applied to different datasets at negligible cost and effort. Under sufficient training, the estimator converges to the Bayesian posterior given the data patterns. As an illustration, we construct a pretrained estimator for a sequential search model (available at pnnehome.github.io). Estimation takes only seconds and achieves high accuracy on 12 real datasets. More broadly, pretrained estimators can make structural models much easier to use and more accessible.