GitHub and Hugging Face host Training Agents IRL: six speakers on RL for agents in SF
ben_burtenshaw · x · 2026-09-29
Hugging Face and GitHub are co-hosting Training Agents IRL, an open-source community evening on October 22 (6–9pm Pacific) at GitHub's SF office, part of Open Source AI Week.
- Format: six 10-minute talks with Q&A, plus a closing panel on what works in RL for agents, the bottlenecks, and what the open-source ecosystem needs next
- Focus is beyond single-turn RL: multi-step, environment-driven training — reward design, rollouts, benchmarks, and the gap between training gains and real-world agent behavior
- Speakers: Govind Kamtamneni (Microsoft), Victor Barres (Mercor), Zach Wentz (Reflection), Tao Lin (RadixArk), Xiangyi Li (Benchflow), and Ben Burtenshaw (HF)
- Attendance requires host approval; the event will also be live-streamed on the HF YouTube channel
Related event: GitHub and Hugging Face to Co-host Training Agents IRL Event(2 posts)→
More from coding & agent
- Who grades the grader? A practical method for testing checks written by coding agents — jonah_omninode · 2026-09-29
- Perplexity launches multi-agent trading, running models like GPT-6 Astra and Grok 4.7 on Coinbase — AravSrinivas · 2026-09-29
- Perplexity details agent safety engineering: 'governance is an engineering problem' — perplexity_ai · 2026-09-29
- OpenRouter coding model share: GLM 5.3 Flash leads at 29.2%, DeepSeek V4.1 Flash at 25.6% — togethercompute · 2026-09-29
- InstaCloud raises $8M seed for agent-native serverless cloud that deploys itself — Shruti_0810 · 2026-09-29
- Routing with small models is 200x faster and 400x cheaper than LLMs, reshaping AI software architecture — Pavan_Belagatti · 2026-09-29