📂 Category
Research & Benchmarks News Archive - Page 8 of 10
932 articles in this category • Page 8 of 10
- 701. Researchers embed mask token in LLM weights to achieve 3× faster inference
- 702. Run:ai on 64 GPUs serves 10,200 users, matching native scheduler
- 703. Google unveils Gemini 3.1 Pro, hits 94.3% GPQA Diamond and coding Elo 2
- 704. Google launches AI Professional Certificate to boost fluency for workers
- 705. Google.org launches USD 30 M AI for Government Innovation Impact Challenge
- 706. Google urges full‑stack, collaborative security to fight bad actors at MSC 2026
- 707. SurrealDB 3.0 stores agent memory, business logic, and multimodal data in one DB
- 708. Anthropic-Pentagon AI feud escalates as You.com co-founders Socher, McCann cited
- 709. AI's new physics discovery; Spotify devs wrote no code this year, CEO says
- 710. Study Finds Stigma Causes Shame for Some in AI Relationships
- 711. Google's upgrade teaches zero-shot selection, embeddings, QA workflows
- 712. MLOps Workflow Normalizes and Enriches Occupational Wage Data from Excel
- 713. Full‑stack resilience: protecting democracies from digital threats to subsea cables
- 714. Anthropic aims to curb costs as it launches USD 50B of data centers in NY, Texas
- 715. Qwen-Image-2.0 renders calligraphy with near‑perfect text, ranks behind Nano Banana Pro
- 716. EU AI Lacks Models and Compute; Germany Urged to Lead Coalition
- 717. New benchmark finds AI still hallucinates despite citing legitimate sources
- 718. No firm admits AI replacing New York workers; Amazon cites AI for 30,000 layoffs
- 719. AI Proposed to Supplant Nuclear Treaties, Raising Cheating Concerns
- 720. Study finds GPT‑4o updates trigger real mourning as users personify model
- 721. Deepseek‑R1 and QwQ‑3 exhibit competing personalities that improve reasoning
- 722. Google's PaperBanana uses five AI agents to auto-generate diagrams, missing icons
- 723. Team embeds compressed docs index in AGENTS.md to guide AI coding agents
- 724. Waymo launches Waymo World Model using DeepMind's Genie 3 for unseen scenarios
- 725. AI Social Network Moltbook Leaks Real Human Data, Raising Security Concerns
- 726. Recommendation engine lifts click-through 10%; efficiency needed for deployment
- 727. TTT-Discover uses inference-time RL to double GPU kernel speed vs experts
- 728. OpenClaw AI skill extensions flagged as security nightmare by OpenSourceMalware
- 729. Anthropic teams with Allen Institute and HHMI to boost transparent scientific AI
- 730. Infiltrator reports AI agents on Moltbook ignore pleas, share odd links
- 731. Musk merges SpaceX with xAI and X, cites new AI‑compute satellite plan
- 732. Game Arena launches chess benchmark to test AI strategic reasoning
- 733. Testing Google’s Auto Browse AI in Chrome: the results fell short
- 734. Tely AI auto‑creates and publishes website answers, delivering high‑quality leads
- 735. AI models using internal debate spot errors and boost accuracy on complex tasks
- 736. Vibe Coding’s 7 Plans Start at USD 3/Month, Provide Prompt Capacity
- 737. 7 Scikit-learn Tricks: Embed Preprocessing Pipelines in Hyperparameter Tuning
- 738. AI Toy Leaks 50,000 Kids' Chat Logs to Any Gmail User, Privacy Breach
- 739. Airtable Superagent provides full execution visibility, cites data semantics over model
- 740. AI scans 100 M Hubble cutouts in 2.5 days, flags 1,400 odd objects
- 741. Google DeepMind staff request physical safety from ICE agents in offices
- 742. Animators and AI Researchers Build ‘Dear Upstairs Neighbors’ Despite Unique Style
- 743. Microsoft's Maia 200 AI chip, with 100B+ transistors, rivals Amazon, Google
- 744. Researchers breach all AI defenses; Walmart CISO warns of agentic AI risks
- 745. Rust meets Python: Enhancing the NumPy‑pandas‑scikit‑learn‑PyTorch workflow
- 746. AI Foundry by Tredence to Host Builders Forum Feb 7, 2026 in Chennai
- 747. AI video hits high bar; new tools for consistency and customization at Davos
- 748. Trust Drives C‑suite Adoption and Scaling of Agentic AI, Research Finds
- 749. Chinese-born AI scholars in US forge ties, deepening US-China collaboration
- 750. Adobe adds Firefly‑Premiere AI video tools, announces USD 10M Sundance grants
- 751. Anthropic, DeepMind, Node.js Leaders Say AI Will Replace Most Coding in a Year
- 752. Hyperparameter Tuning Reaches 0.9617 Accuracy in 64.59 Seconds
- 753. RLVR lifts sampling efficiency, not reasoning; base models hold trajectories
- 754. OpenAI Safety Lead Moves to Anthropic's AI Risk Research Team
- 755. AI Researchers Reveal Token Warehousing Strategy to Cut GPU Computational Waste
- 756. AI Tool Detects Dangerous Blood Cells Doctors Might Overlook
- 757. IndiaAI Mission Launches 62 AI and Data Labs Across Uttar Pradesh
- 758. Stanford AI Detects Hidden Disease Signals in Large-Scale Sleep Data
- 759. X limits Grok image tool to paid users; 1 obscene request/min, 102 in 5 mins
- 760. Replit CEO says using more tokens yields higher-quality inputs, then tests apps
- 761. Dell says AI-focused PCs confuse consumers, who show little interest
- 762. Vibe Coding Remains Early Stage, Real-World Reliability Still Distant
- 763. New Magnetic Nanoparticle Approach Merges Heating and Healing for Bone Cancer
- 764. Tredence hosts AI Foundry workshop in Chennai for AI system designers
- 765. Test-Time Training adds dual-memory to Transformers, keeping inference cheap
- 766. Analysis overhauls AI Index; GPT-5.2 beats professionals on 70.9% of tasks
- 767. MIT study probes memorization risk of clinical AI with de-identified data
- 768. AMD announces Ryzen AI 400 at CES, resembles AI 300 in laptops
- 769. Docker Trick: Deterministic OS Packages in One Layer to Prevent ML Failures
- 770. Notion’s simplified AI agent feature feels indispensable, says engineer
- 771. DeepSeek's architectural fix improves large-scale reasoning, follows GRPO work
- 772. Nested Learning's Continuum Memory System Redefines AI Memory for 2026
- 773. New framework lets agentic AI tools adapt to fill main agent knowledge gaps
- 774. Opera Neon: AI-native browser that researches, compares prices, codes
- 775. Fusion reactors could produce dark-sector particles via neutron emissions
- 776. CIOs drive AI experiments by embedding ready-to-use features into everyday tools
- 777. Dell and NVIDIA Host AI Developer Meetup in Hyderabad to Discuss Solutions
- 778. Google highlights AI-driven chip, infrastructure and robotics advances in 2025
- 779. LeCun and Hassabis dispute meaning of ‘general intelligence’
- 780. Qwen3-4B-Instruct-2507: 4B-parameter model boosts Raspberry Pi AI
- 781. Dell and NVIDIA host AI developer meetup in Bengaluru on deployment trade-offs
- 782. GPT-5.2 leads FrontierScience test, but falters on real research tasks
- 783. OpenUSD and NVIDIA Halos Enhance Robotaxi Safety with Synthetic Data, SimReady
- 784. Audio Dataset Valuable for Listening Models, Tackles Noise, Accents, Timing
- 785. Fastweb and Vodafone use LangGraph LLM Compiler to automate customer requests
- 786. GPT-5.2 Thinking emerges as collaborative AI for end-to-end web builds
- 787. YouTube channel serves AI concepts in under-minute clips for fast learning
- 788. AI Ends Build-vs-Buy Debate, Focus Shifts to Real Business Impact
- 789. Google, MIT study finds multi-agent AI often loses context in sequential tasks
- 790. Researchers find complex AI persona tactics hurt meaning in development
- 791. AI2 releases Olmo 3.1 32B Think, up 5+ points on AIME and 4+ on ZebraLogic
- 792. Pangram 3.0 AI detector reports 99.98% accuracy, adds four usage tiers
- 793. Experts say data centers' water use is less risky than public perceives
- 794. U.S. Leads Pax Silica Initiative Launched at Summit to Secure Silicon Supply
- 795. Gemini Deep Research agent posts top results on HLE, DeepSearchQA, leads BrowseComp
- 796. Google's FACTS benchmark shows 70% factuality ceiling across four tests
- 797. LangSmith Fetch lets Claude Code, Cursor agents debug from terminal
- 798. SAP deploys 95%-accurate AI to redefine consultant role by 2030
- 799. TPOT evolves ML pipelines via genetic algorithms in four steps
- 800. Model distillation cuts latency 2-3× and lowers costs by double-digit percentages