Interconnects

84 Episodes

1. GPT-4.5: "Not a frontier model"?

   Published: 2/28/2025
2. Character training: Understanding and crafting a language model's personality

   Published: 2/26/2025
3. Claude 3.7 thonks and what's next for inference-time scaling

   Published: 2/24/2025
4. Grok 3 and an accelerating AI roadmap

   Published: 2/18/2025
5. An unexpected RL Renaissance

   Published: 2/13/2025
6. Deep Research, information vs. insight, and the nature of science

   Published: 2/12/2025
7. Making the U.S. the home for open-source AI

   Published: 2/5/2025
8. Why reasoning models will generalize

   Published: 1/28/2025
9. Interviewing OLMo 2 leads: Open secrets of training language models

   Published: 1/22/2025
10. DeepSeek R1's recipe to replicate o1 and the future of reasoning LMs

    Published: 1/21/2025
11. Let me use my local LMs on Meta Ray-Bans

    Published: 1/15/2025
12. (Voiceover) DeepSeek V3 and the actual cost of training frontier AI models

    Published: 1/9/2025
13. The state of post-training in 2025

    Published: 1/8/2025
14. Quick recap on the state of reasoning

    Published: 1/2/2025
15. (Voiceover) 2024 Interconnects year in review

    Published: 12/31/2024
16. (Voiceover) OpenAI's o3: The grand finale of AI in 2024

    Published: 12/20/2024
17. (Voiceover) The AI agent spectrum

    Published: 12/18/2024
18. (Voiceover) OpenAI's Reinforcement Finetuning and RL for the masses

    Published: 12/11/2024
19. Interviewing Finbarr Timbers on the "We are So Back" Era of Reinforcement Learning

    Published: 12/5/2024
20. (Voiceover) OpenAI's o1 using "search" was a PSYOP

    Published: 12/4/2024