ResearchCoordinated navigation reduces urban congestion: a large-scale routing-app experimentA six-month field experiment in 10 major US cities, described in a Nature Cities paper, tested a modified Google Maps routing that gently redirects a small share of trips away from recurring bottlenecks.4 min read
ResearchAnthropic finds a self-organized 'J-space' in Claude that hosts hidden internal reasoningAnthropic researchers report the emergence of a small internal zone in the Claude model’s neural network, dubbed the J-space, which encodes concepts the model can reason about without necessarily verbalizing them.4 min read
ResearchFixed Compute Budgets Skew AI Benchmark Results, UK Study FindsThe UK's AI Security Institute tested frontier models on seven benchmarks and found that fixed compute budgets baked into evaluations systematically understate model capabilities.2 min read
ResearchAnthropic reveals a distinct internal 'J‑Space' in Claude that holds off‑line computationsAnthropic announced it has identified a small internal workspace inside its Claude model, called "J‑Space," where the model can activate concepts and run computations separate from the outputs it shares with users.3 min read
ResearchNonuniform Tensor Parallelism (NTP) to Preserve Goodput in Large‑Scale LLM TrainingNonuniform Tensor Parallelism (NTP) is an experimental framework that keeps large, tightly coupled LLM training jobs productive when some GPUs become transiently unavailable.5 min read
ResearchAnthropic finds a ‘workspace’ inside Claude resembling global workspace theoryAnthropic researchers report that Claude language models developed a small, reportable internal workspace — a “J-space” — revealed by a new Jacobian-based interpretability method.6 min read
ResearchNeuroscience, philosophy and interpretability experts evaluate the researchThe research team invited experts in neuroscience, philosophy and interpretability to review their work and share their perspectives.1 min read
ResearchAnthropic: discovery of global-workspace-like operation in the Claude language modelAnthropic's new research reports that the Claude language model exhibits a 'global workspace'-like structure similar to the conscious and unconscious partitioning observed in the brain, in which only a small fraction of internal states become explicit.1 min read
ResearchOSWORLD 2.0: a benchmark for long‑horizon AI computer useOSWORLD 2.0 is a new benchmark, developed by a consortium of universities and industry partners, that measures how well AI systems perform multi‑step, multi‑program tasks on real computer environments.3 min read
ResearchRemote Labor Index shows rapid improvement in AI handling of high-value online freelance tasksResearchers at the Center for AI Safety and Scale Labs report that AI systems' success on the Remote Labor Index rose from 2.5% in October 2025 to 16.1% in July 2026, indicating much faster capability growth on economically valuable online projects.3 min read
ResearchSemmelweis University opens second German campus in partnership with Westpfalz-KlinikumSemmelweis University has signed a ten-year strategic agreement with Westpfalz-Klinikum in Kaiserslautern to establish the university’s second campus in Germany.2 min read
ResearchResearchers: rest periods similar to human sleep could be useful for large language modelsResearchers at Carnegie Mellon Egyetem and Marylandi Egyetem say that not only humans need sleep for cognitive organization, but large language models (LLM) could also benefit from rest periods that mimic human sleep patterns.1 min read
