Uploaded July 2026 | Updated September 2026, 2 weeks ago

---

Jack Min Ong from Prime Intellect Inc will discuss continual learning for long-running agents and how adaptive systems can improve over time through feedback, evaluation, and responsible release practices.

Key Takeaways:
Continual learning can help agents adapt to new information and improve long-running workflows
Feedback loops need evaluation, governance, and clear boundaries before they influence production behavior
Adaptive agents require measurement and validation so improvements remain trustworthy and useful

Industry: All Industries
Topic: Agentic AI and Reasoning AI,Claws & Long Running Agents / Generative AI - Text Generation
Technical Level: Technical - Intermediate
Intended Audience: Developer / Engineer
NVIDIA Technology: Blackwell,Cloud / Data Center GPU,cuBLAS,CUDA,cuDDN,Ethernet Networking,Grace CPU,HGX,Hopper,Infiniband Networking,Interconnect Networking,NCCL,NeMo,NVLink / NVSwitch,RTX GPU

Build, monitor, and optimize AI agents with NVIDIA NeMo: nvidia.com/en-us/ai-data-science/products/nemo
Explore NVIDIA Nemotron models: nvidia.com/en-us/ai-data-science/foundation-models/nemotron

#NVIDIAGTC