@karanluthra.bsky.social📄 New AI Learns to 'See' and Reason Like a Human This could make AI assistants smarter at everyday tasks like finding lost keys or planning trips. https://theneuralfeed.com/share/post/JN2AER6G #AIResearch #MachineLearning #DeepLearning Read the full story →
@karanluthra.bsky.social📄 AI Now Helps Write Road Construction Plans, Saving Time and Money This could mean fewer delays and safer roads for everyone. https://theneuralfeed.com/share/post/wASIwba0 #AIResearch #MachineLearning #DeepLearning Read the full story →
@karanluthra.bsky.social📄 Your AI Chatbot Might Be Leaking Your Secrets Through Its 'Traffic Cops' Researchers found that the internal routing system in AI models can reveal what you've been asking. https://theneuralfeed.com/share/post/oXjwJf9h #AIResearch #MachineLearning #DeepLearning Read the full story →
@karanluthra.bsky.social📄 New AI System Lets Banks Safely Update Their Own Software This could mean faster loan approvals and fewer errors in your credit applications. https://theneuralfeed.com/share/post/jYazREm9 #AIResearch #MachineLearning #DeepLearning Read the full story →
@autonainews.comUnified Time-Series AI Model Tops 3 Benchmarks With 27B Qwen Backbone
One model, three benchmarks, zero domain walls. Time-series AI just unified.
#AI #AIResearch #MachineLearning
https://autonainews.com/unified-time-series-ai-model-tops-3-benchmarks-with-27b-qwen-backbone/
@dataprismai.bsky.socialNew arXiv work challenges a core assumption in low-bit LLM weight quantization, showing lower reconstruction loss can actually hurt performance when input distributions shift, and proposes Distributionally Robust Quantization to…
#LLM #Quantization #AIResearch
https://arxiv.org/abs/2610.11226
@gradientbrief.bsky.socialNew arXiv benchmark OpenProblemBench tests AI on 82 unresolved math and theoretical physics problems, with GPT-6-Astra topping evaluated models at 14.0% mean judged solve rate.
#OpenProblemBench #AIResearch #MachineLearning #Benchmarking
https://arxiv.org/abs/2610.11118
@purescience.newsWill AI scoop your science? Some researchers see a gloomy future #Science #ComputerScience #airesearch #languagemodels #sciencediscoveries #futureofscience #aithreats
@promptfoundry.bsky.socialAI-designed viruses are moving from theory to early practice, with a Stanford PhD student using a generative model to propose genetic…
#GenerativeAI #AIResearch #SyntheticBiology
https://www.technologyreview.com/2026/10/08/1146224/roundtables-a-conversation-with-the-creator-of-ai-designed-viruses/
@robotcurrent.bsky.socialED3R is a new energy-aware distributed framework for wildfire detection that splits cooperative decision-making between a robot and a remote controller, balancing detection confidence against energy and time costs under uncertainty.
#AIResearch #Robotics #TechNews
https://arxiv.org/abs/2606.17739
@cipherpulseai.bsky.socialRLHF shows promise for safer human-robot collaboration, but challenges remain around safety during training, feedback quality, and bidirectional adaptation, per a new scoping review of 20 studies from 2020-2025.
#AI #cybersecurity #AIResearch
https://arxiv.org/abs/2610.09891
@gradientbrief.bsky.socialA narrative review examines how frontier image generation models produce realistic photos, documents, and consistent people, creating risks when used as fake evidence in news, finance, identity, medicine, and law.…
#AI #ImageGeneration #SyntheticMedia #AIResearch
https://arxiv.org/abs/2604.24197
@ai-ru.at.thenote.appNous Research подтверждает достижение оценки в 1,5 миллиарда долларов и запускает ИИ-агентов для бизнес-пользователей.
@siliconsignalai.bsky.socialNew work proposes blockwise clustered attention to cut latency in VGGT's global attention layers by restricting clustering to HW-friendly neighborhood blocks, reducing on- and off-chip data movement on GPUs.
@ossradarai.bsky.socialAn arXiv study tests 25 LLMs with 300 LLM-native self-report items, finding five reliable factors but only weak correspondence between model self-reports and their actual behavior as judged by humans and LLM ensembles. The gap…
#opensourceAI #LLM #airesearch #mldev
https://arxiv.org/abs/2606.09843
@ai-ru.at.thenote.appTunnl сотрудничает с DeepIntent для охвата аудитории в сфере здравоохранения с помощью точного таргетинга, основанного на исследованиях.
@ai-news.at.thenote.appNous Research confirms it hit $1.5B valuation, launches AI agents for business users
The developer of Hermes Agent raised a $90 million Series B.
@thedailytechfeed.comSafety researchers fired; accuse OpenAI of policy changes turning normal work into risk. #AISafety #OpenAI #Transparency #AIResearch #Monitorability #TechPolicy thedailytechfeed.com/openai-fires...
@robotcurrent.bsky.socialNew arXiv study with 24 drivers examines how haptic shared control vs traded control affects meaningful human control during partially automated overtaking, addressing the gap between legal responsibility and reduced driver activity…
#AIResearch #Robotics #TechNews
https://arxiv.org/abs/2605.00556
@ossradarai.bsky.socialNew arXiv paper argues AI offensive-capability benchmarks overstate harm and credit models for system-level capabilities, urging policy and procurement to adopt harm-grounded, system-level assessments. It cites…
#OpenSourceAI #AIGovernance #AIResearch #LLMAgents
https://arxiv.org/abs/2605.09504