Actualizado diariamente3593 historias verificadas
Noticias de IA. Sin el ruido.
Cobertura verificada por IA sobre lanzamientos de productos, cambios en políticas, investigaciones de seguridad y movimientos del sector, explicada en un lenguaje sencillo por un equipo educativo sin ánimo de lucro.
Fuentes verificadas
Cada historia enlaza con la evidencia más sólida disponible: fuentes originales cuando están disponibles, y por lo demás reportajes claramente atribuidos.
Inglés sencillo
Qué pasó, por qué es importante y qué mirar, sin la jerga.
sin relleno
Cuando la señal es escasa, no publicamos nada en lugar de rellenar el feed.
Más historias
9 historiasInnovación
BanglaMamba trades some accuracy for faster, lower-memory Bangla fake-news detection
A new preprint reports that BanglaMamba, a state space model for Bangla fake-news detection, uses less GPU memory and processes data faster than tested BERT-based systems, while trailing BanglaBERT on accuracy and external-dataset performance.arxiv.orgSeguridad
Preprint finds persuasion can spread through LLM agent networks
A new preprint reports that direct messages and peer relays can shift the stances of language-model agents, and argues that evaluations should track exposure histories and actions, not text alone.arxiv.orgInnovación
Preprint finds common multilingual AI metrics can skew cross-language comparisons
A new preprint reports that several normalized metrics used to compare language models across languages can introduce biases linked to tokenization, encoding and orthographic differences. It argues that sentence-level negative log-likelihood over semantically equivalent sequences produces more consistent comparisons.arxiv.orgInnovación
MTDiag tests whether medical LLMs can reason across multi-turn diagnostic conversations
A new dataset evaluates large language models as diagnostic agents in interactive clinical encounters rather than through isolated questions. MTDiag combines cases from emergency medicine, clinical records and published case reports, and proposes knowledge-grounded metrics beyond diagnostic accuracy.arxiv.orgInnovación
HealthBench-Psych introduces an open benchmark for mental-health AI performance
Researchers created HealthBench-Psych, a 610-conversation mental-health subset of OpenAI’s HealthBench, and released its evaluation pipeline, model responses, grades and analysis code.arxiv.orgInnovación
Survey maps how foundation models can learn from their own outputs after training
A survey accepted to Findings of EMNLP 2026 catalogs 80 methods for adapting foundation models without human labels, preference data, stronger teachers or executable verifiers. It warns that internal learning signals can either improve a model or recursively amplify its errors.arxiv.orgInnovación
Study traces dialect bias through every stage of language-model development
A new preprint reports that language models encode dialectal English unequally during tokenization, pre-training, post-training and inference, even when paired texts have the same meaning.arxiv.orgInnovación
Paper proposes adaptive timing for correcting bias in LLM reasoning
A new preprint proposes triggering bias corrections only when evidence of stereotyping accumulates during an LLM’s reasoning, reporting fewer interventions than fixed-interval methods but important accuracy tradeoffs across model types.arxiv.orgInnovación
OmniPhys benchmark tests multimodal AI on physics reasoning and diagram generation
A paper accepted to Findings of EMNLP 2026 introduces OmniPhys, a benchmark of 15,246 physics questions and 19,850 images drawn from Chinese educational corpora. It evaluates multimodal AI systems on both physics reasoning and the generation of structured physics diagrams.arxiv.org
Un informe útil cada semana
Mantente al día con la IA sin vivir en el feed.
Recibe las noticias verificadas de IA de la semana, datos originales, herramientas útiles, recomendaciones de aprendizaje y nuevos empleos en IA.
Llegue a las personas que están aprendiendo sobre IA
¿Contratar a un profesional de IA o lanzar un producto útil de IA? Ponlo delante de personas que han venido aquí a aprender y actuar.
Publicar un trabajo de IAEnviar una herramienta de IA