Search articles by title, topic or section.
Khabre desk
Recent reporting, analysis and explainers from the large-language-models desk.
health
A blinded preprint compared five large language models on 17 synthetic emergency-department cases. All named the reference diagnosis, but triage assignments and clinician ratings of response quality differed.
health
A medRxiv preprint tested eight general-purpose large language models on 100 neurology board-style questions. Confidence estimates were often too high, while agreement across repeated answers best distinguished correct from incorrect responses.
health
A medRxiv preprint introduces KampoBench, a physician-revised benchmark testing 10 large language models on 74 Japanese Kampo consultation scenarios. The results point to a gap between naming relevant information and applying it to a specific case.
technology
A randomized experiment at ICML 2026 found near-zero effects of LLM-use policies on paper decisions, scores and reviewer confidence. In a post-study survey, reviewers reported substantial noncompliance with both restrictive and permissive rules.
health
A medRxiv preprint reports that a RoBERTa model classified depression-related Twitter users with 0.822 accuracy on an 80/20 training–testing split. The study also used GPT-4o to identify themes in posts, including societal pressure and low self-esteem.
technology
Anthropic has introduced Claude Opus 5.5, the first model in its Claude 5.5 family. The company says it improves coding, knowledge work and safety performance while costing 40% less than Opus 5 on typical workloads.
health
A medRxiv preprint evaluated 14 general-purpose large language models on eight schizophrenia symptoms in clinical summaries from a 704-person severe mental illness cohort. Performance was poor to moderate overall, with negative symptoms extracted less accurately than positive symptoms.
science
A bioRxiv preprint introduces MindEvolve, an autonomous workflow that uses large language models to generate interpretable symbolic models of human decision-making in social interactions. The models performed most reliably on simpler economic and social preferences, while some advanced systems showed promise on theory of mind and recursive planning.
health
A medRxiv preprint evaluated a large-language-model pipeline for assessing risk of bias in neurological prognosis studies. Agreement with original human assessments was limited, while a small comparison suggested a similar level of agreement to that between human raters.
technology
Anthropic says Claude Fable 5.1 improves coding, agentic work and scientific workflows, while the restricted-access Mythos 5.1 uses the same underlying model with different safeguards.