--- schema: 1 type: video title: 'Modern Post-Training: A Deep Dive — Will Brown, Prime Intellect' description: >- Deep dive into Prime Intellect's open-source ecosystem of post-training tools, including the verifiers and prime-rl libraries, as well as the Lab platform for self-serve training and inference. Speaker: Will Brown — Research Lead, Prime Intellect Will Brown leads Applied Research at Prime Intellect and builds open research infrastructure to enable every company to train, deploy, and self-improve their own frontier agentic models. He holds a PhD in Computer Science from Columbia University. X: https://x.com/willccbb LinkedIn: https://www.linkedin.com/in/willcb/ GitHub: https://github.com/willccbb Website: https://willcb.com TImestamps 0:00 Introduction and Overview of Prime Intellect 4:20 Defining the Environment in Post-Training 9:33 Decomposing Environments: Tasks, Harnesses, and Runtimes 12:46 Verifiers V1: The New Modular Pattern 17:46 Rewards, Metrics, and Group-Level Rewards 20:25 Tooling, User Simulators, and MCP Integration 22:00 The Interception Server Pattern 24:13 Trace Graphs and Handling Tokenization 25:35 The Renderers Library for Chat Templates 29:20 Primaril: Asynchronous Reinforcement Learning 38:02 Customizing Training Algorithms and Losses 42:35 The Lab Platform and Hosted Training --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/V-EDrhIhHzQ-modern-post-training-a-deep-dive-will-brown-prime-intellect Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- canonical: >- https://liquidgold.cc/channel/aidotengineer/explore/video/modern-post-training-a-deep-dive-will-brown-prime-i-1e0e7d6cbbe1 markdown: >- https://liquidgold.cc/channel/aidotengineer/video/modern-post-training-a-deep-dive-will-brown-prime-i-1e0e7d6cbbe1.md llms: null parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T07:51:15.922Z' language: null children: [] videoId: V-EDrhIhHzQ durationSeconds: 2812 publishedAt: '2026-07-13T15:34:07.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 273 endSeconds: 380 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=273' heldReason: rights - ordinal: 1 text: null startSeconds: 390 endSeconds: 509 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=390' heldReason: rights - ordinal: 2 text: null startSeconds: 390 endSeconds: 509 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=390' heldReason: rights - ordinal: 3 text: null startSeconds: 520 endSeconds: 575 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=520' heldReason: rights - ordinal: 4 text: null startSeconds: 586 endSeconds: 771 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=586' heldReason: rights - ordinal: 5 text: null startSeconds: 784 endSeconds: 1066 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=784' heldReason: rights - ordinal: 6 text: null startSeconds: 784 endSeconds: 1066 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=784' heldReason: rights - ordinal: 7 text: null startSeconds: 1076 endSeconds: 1231 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1076' heldReason: rights - ordinal: 8 text: null startSeconds: 1076 endSeconds: 1231 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1076' heldReason: rights - ordinal: 9 text: null startSeconds: 1241 endSeconds: 1327 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1241' heldReason: rights - ordinal: 10 text: null startSeconds: 1337 endSeconds: 1459 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1337' heldReason: rights - ordinal: 11 text: null startSeconds: 1337 endSeconds: 1459 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1337' heldReason: rights - ordinal: 12 text: null startSeconds: 1471 endSeconds: 1538 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1471' heldReason: rights - ordinal: 13 text: null startSeconds: 1549 endSeconds: 1760 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1549' heldReason: rights - ordinal: 14 text: null startSeconds: 1771 endSeconds: 1868 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1771' heldReason: rights - ordinal: 15 text: null startSeconds: 1879 endSeconds: 1978 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1879' heldReason: rights - ordinal: 16 text: null startSeconds: 1988 endSeconds: 2112 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=1988' heldReason: rights - ordinal: 17 text: null startSeconds: 2123 endSeconds: 2211 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=2123' heldReason: rights - ordinal: 18 text: null startSeconds: 2223 endSeconds: 2288 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=2223' heldReason: rights - ordinal: 19 text: null startSeconds: 2298 endSeconds: 2561 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=2298' heldReason: rights - ordinal: 20 text: null startSeconds: 2298 endSeconds: 2561 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=2298' heldReason: rights - ordinal: 21 text: null startSeconds: 2572 endSeconds: 2812 watch: 'https://liquidgold.cc/watch/V-EDrhIhHzQ?t=2572' heldReason: rights --- # Modern Post-Training: A Deep Dive — Will Brown, Prime Intellect Deep dive into Prime Intellect's open-source ecosystem of post-training tools, including the verifiers and prime-rl libraries, as well as the Lab platform for self-serve training and inference. Speaker: Will Brown — Research Lead, Prime Intellect Will Brown leads Applied Research at Prime Intellect and builds open research infrastructure to enable every company to train, deploy, and self-improve their own frontier agentic models. He holds a PhD in Computer Science from Columbia University. X: https://x.com/willccbb LinkedIn: https://www.linkedin.com/in/willcb/ GitHub: https://github.com/willccbb Website: https://willcb.com TImestamps 0:00 Introduction and Overview of Prime Intellect 4:20 Defining the Environment in Post-Training 9:33 Decomposing Environments: Tasks, Harnesses, and Runtimes 12:46 Verifiers V1: The New Modular Pattern 17:46 Rewards, Metrics, and Group-Level Rewards 20:25 Tooling, User Simulators, and MCP Integration 22:00 The Interception Server Pattern 24:13 Trace Graphs and Handling Tokenization 25:35 The Renderers Library for Chat Templates 29:20 Primaril: Asynchronous Reinforcement Learning 38:02 Customizing Training Algorithms and Losses 42:35 The Lab Platform and Hosted Training --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/V-EDrhIhHzQ-modern-post-training-a-deep-dive-will-brown-prime-intellect Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- Canonical: https://liquidgold.cc/channel/aidotengineer/explore/video/modern-post-training-a-deep-dive-will-brown-prime-i-1e0e7d6cbbe1 · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.