Agent performance can look great in training. Then it meets the real world.
Coming to Open Source AI Week: Training Agents IRL, hosted by
@github and
@huggingface.
We’re going beyond single-turn RL to explore how agents learn and act across multi-step environments:
🎯 Reward design
🔁 Rollouts and environment-driven training
📊 Benchmarks and evaluation
🌍 The gap between training gains and real-world behavior
Hear short talks from researchers and builders at Hugging Face, Microsoft, Mercor, Reflection, RadixArk, and Benchflow—followed by a panel on what works, where the bottlenecks are, and what the open-source ecosystem needs next.
📅 October 22, 2026 · 6–9 PM
📍 GitHub · San Francisco
Request your spot →
#
OpenSourceAIWeek# #
OpenIntelligence#