Interconnects

78 Episoade

1. Llama 3.1 405b, Meta's AI strategy, and the new open frontier model ecosystem

   Publicat: 23.07.2024
2. SB 1047, AI regulation, and unlikely allies for open models

   Publicat: 17.07.2024
3. Switched to Claude 3.5

   Publicat: 03.07.2024
4. Interviewing Dean Ball on AI policy: CA SB 1047, upcoming AI disaster response, Llama 3 405B, Chinese open-source AI, and scaling laws

   Publicat: 27.06.2024
5. RLHF Roundup: Trying to get good at PPO, charting RLHF's impact, RewardBench retrospective, and a reward model competition

   Publicat: 26.06.2024
6. Frontiers in synthetic data

   Publicat: 21.06.2024
7. Text-to-video AI is already abundant

   Publicat: 18.06.2024
8. AI for the rest of us

   Publicat: 12.06.2024
9. A realistic path to robotic foundation models

   Publicat: 05.06.2024
10. We aren't running out of training data, we are running out of open training data

    Publicat: 29.05.2024
11. Name, image, and AI's likeness

    Publicat: 22.05.2024
12. OpenAI chases Her

    Publicat: 16.05.2024
13. OpenAI's Model (behavior) Spec, RLHF transparency, and personalization questions

    Publicat: 13.05.2024
14. RLHF: A thin line between useful and lobotomized

    Publicat: 01.05.2024
15. Phi 3 and Arctic: Outlier LMs are hints

    Publicat: 30.04.2024
16. AGI is what you want it to be

    Publicat: 24.04.2024
17. Llama 3: Scaling open LLMs to AGI

    Publicat: 21.04.2024
18. Stop "reinventing" everything to "solve" alignment

    Publicat: 17.04.2024
19. The end of the "best open LLM"

    Publicat: 15.04.2024
20. Why we disagree on what open-source AI should be

    Publicat: 03.04.2024