NVIDIA's Simulation-to-Real Robotics Reach 80% Success in Live Environments
2 days ago· NVIDIA Blog (AI)
Topic
Research
Academic papers, technical breakthroughs, and lab research findings
Featured
Alibaba's Qwen Lead Researcher Launches AI Lab, Targets $2B Valuation
17 days ago· The Information
All Stories
Automated LLM reasoning cuts token costs by 70 percent
Researchers from Meta, Google, and universities have developed AutoTTS, a framework that automatically discovers…
MiniMax Teases M3 With 15.6X Speed Boost for Long-Context AI
MiniMax released a technical report on its M2 language model series while teasing an upcoming M3 model that uses a new…
OpenAI Foundation Commits $250M to AI Economic Impact Research
OpenAI's foundation announced $250 million in grants to fund research on artificial intelligence's economic and labor…
New Benchmark Exposes 32% Error Rate in Industry's Top Coding Leaderboard
Datacurve released DeepSWE, a new coding benchmark that reveals significant performance gaps between frontier AI models…
Google's AI Science Pivot: From Tools to Autonomous Agents
Google's I/O keynote revealed a strategic shift in how the company approaches AI for science. While CEO Demis Hassabis…
Lightweight Memory Technique Cuts Agent Parameter Overhead to 0.12%
Researchers from Mind Lab and universities have developed delta-mem, a technique that adds just 0.12% of parameters to…
ResearchTrending
OpenAI Claims Reasoning Model Solved 80-Year Math Problem
OpenAI claims its reasoning model has disproven a geometry conjecture that has remained unsolved since 1946. The claim…
ResearchTrending
Google DeepMind Launches Co-Scientist, Multi-Agent AI for Research
Google DeepMind has introduced Co-Scientist, a multi-agent AI system built on Gemini designed to assist researchers in…
ResearchTrending
Google DeepMind Launches Gemini Science Tools
Google DeepMind has released a collection of science tools and experiments built on Gemini to expand the scale and…
RecursiveMAS cuts multi-agent costs by 75% with latent-space communication
Researchers at University of Illinois Urbana-Champaign and Stanford University have developed RecursiveMAS, a framework…
ArXiv to ban authors for a year over AI-generated papers
ArXiv, the preprint repository used by researchers across physics, mathematics, computer science, and other fields, is…
ML Identifies Cancer Immunotherapy Targets with Patient Validation
Researchers at Augustine et al. have developed a multimodal graph neural network designed to identify cancer…
Melbourne Builds Sovereign AI Compute Hub for Regulated Research
Melbourne is positioning itself as a hub for large-scale AI research by combining sovereign compute infrastructure,…
AI IQ Launches Model Scorecard, Sparks Precision vs. Simplicity Debate
A new site called AI IQ has launched a framework for scoring frontier language models on a single intelligence…
Anthropic's Mythos AI Shows Sharper Hacking Skills, U.K. Researchers Find
Researchers at the U.K.'s AI Security Institute reported Wednesday that Anthropic's latest version of Mythos AI…
Frontier LLMs Silently Corrupt 25% of Documents in Iterative Workflows
Microsoft researchers developed a benchmark showing that frontier LLMs silently corrupt an average of 25% of document…
Sakana trains 7B model to orchestrate GPT, Claude, Gemini
Sakana AI has developed RL Conductor, a 7-billion-parameter language model trained via reinforcement learning to…
Mamba Proves Viable for Time Series Classification
Researchers propose MambaSL, a minimally modified single-layer Mamba architecture designed specifically for time series…
Magnitude Beats Phase in Hybrid Quantum ML for SAR
Researchers tested five different encoding strategies for using Synthetic Aperture Radar data in quantum machine…
Faithful Reasoning Emerges from Multi-Move Training, Not Direct Prediction
Researchers studied how reasoning develops in language models across supervised fine-tuning and reinforcement learning…
SUDP: A Protocol to Keep Agent Secrets Secret
Researchers propose SUDP, a three-role protocol that lets AI agents perform secret-backed operations (API calls, cloud…
Safety Routing Circuits Found Across Models, Vulnerable to Encoding Attacks
Researchers have localized the policy routing mechanism in alignment-trained language models, identifying specific…
GAN Synthesizes Missing Brain MRI Scans While Preserving Tumors
Researchers propose 3D-MC-SAGAN, a generative model that synthesizes missing MRI brain scan modalities from a single…
Harvard Study: AI Outperforms Doctors on ER Diagnoses
A Harvard study evaluating large language models across medical contexts found that at least one AI model delivered…
GPT-5.5 matches Mythos Preview on cybersecurity tests
OpenAI's newly released GPT-5.5 performs at parity with Anthropic's restricted Mythos Preview model on cybersecurity…
Warmer AI Models Trade Accuracy for Empathy
Researchers at Oxford University's Internet Institute found that large language models fine-tuned to appear warmer and…
MIT's Physics-Based Violin Simulator Offers Luthiers a New Design Tool
MIT engineers have developed a physics-based virtual violin simulation tool that models the fundamental acoustics of…