Agora - The Marketplace of Ideas

The Bleeding Edge

38 min · 16 de nov de 2025
Portada del episodio The Bleeding Edge

Descripción

**Are the state-of-the-art autoregressive decoder-style transformers the only future for large language models?** **We dive into the most fascinating alternatives, including linear attention hybrids that promise huge efficiency gains for long contexts and text diffusion models that generate tokens in parallel instead of sequentially.** Plus, discover how Code World Models are training models to simulate code behavior for improved modeling performance, aiming to develop more capable coding systems.

Comentarios

0

Sé la primera persona en comentar

¡Regístrate ahora y únete a la comunidad de Agora - The Marketplace of Ideas!

Empezar

2 meses por 1 €

Después 4,99 € / mes · Cancela cuando quieras

  • Podcasts exclusivos
  • 20 horas de audiolibros / mes
  • Podcast gratuitos

Todos los episodios

101 episodios