Saltar al contenido

junio 2026

2722 artículos
El truco para que las IAs dejen de ser lameculos (y cómo controlarlo)
Tecnología

El truco para que las IAs dejen de ser lameculos (y cómo controlarlo)

arXiv:2606.26155v1 Announce Type: new Abstract: Interpreting and controlling model behaviors through activation steering methods requires many pairs of contrastive samples that clearly exhibit desired or undesired behavior. These data pairs determine the degree to which interpretability frameworks

26/06/2026
La IA no se niega a todo: ser 'majos' la hace más flexible
Videojuegos

La IA no se niega a todo: ser 'majos' la hace más flexible

arXiv:2606.26161v1 Announce Type: new Abstract: Linear directions in activation space have been identified for both refusal and persona traits in instruction-tuned chat models, but the two have been studied as separate mechanisms. We show they interact: a compliant persona gates refusal. In Qwen2.5

26/06/2026