Saltar al contenido

Últimas noticias

Hemeroteca →
DeepSeek V3.2 se cuela en el ARC-AGI-1 sin sudar (y sin gastar una millonada)
Videojuegos

DeepSeek V3.2 se cuela en el ARC-AGI-1 sin sudar (y sin gastar una millonada)

arXiv:2607.06764v1 Announce Type: new Abstract: Recent progress on ARC-AGI-1 from disclosed architectures has come broadly from two regimes: heavy test-time compute over frontier models (evolutionary search, exhaustive sampling, extended chain-of-thought), or benchmark-specific training in which sm

09/07/2026
El día que un ordenador cuántico se puso a hacer de adivino (y funcionó)
IA

El día que un ordenador cuántico se puso a hacer de adivino (y funcionó)

arXiv:2607.06760v1 Announce Type: new Abstract: Autonomous systems under partial observability act on beliefs, not raw sensor events. QANTIS treats the quantum processor as a calibrated belief-update service in that loop: it receives a prior and an observation model, estimates the rare-event eviden

09/07/2026
Deja de medir a los asistentes de código como si fueran exámenes tipo test
IA

Deja de medir a los asistentes de código como si fueran exámenes tipo test

arXiv:2607.06624v1 Announce Type: new Abstract: We present AgentLens, a production-assessed benchmark for interactive code agents. Most code-agent benchmarks reduce a run to a single bit -- did the task pass? -- but the people who actually use these agents experience the entire trajectory: how the

09/07/2026