NVIDIA AVO Reaches 100% on ARC-AGI-3, Demonstrating a Frontier-Level General-Purpose Architecture for Long-Horizon Autonomous Agents

Summary

NVIDIA’s Agentic Variation Operators (AVO) is a general-purpose agent architecture designed to make frontier language models reliable on extended, multistep tasks. It combines tool use and execution with persistent memory, an iterative inspect-plan-implement-evaluate loop, and a supervisor that can redirect work when progress stalls. In a seven-day GPU-kernel optimization study, AVO explored more than 500 directions, produced 40 committed versions, and outperformed cuDNN by up to 3.5% and FlashAttention-4 by up to 10.5% across evaluated configurations. On the ARC-AGI-3 public set, the same architecture achieved a 100.00 RHAE score across 25 environments, completing all 183 levels in 6,624 actions; the post cautions that comparisons with other systems are not controlled ablations.

Takeaways

1. 1

   AVO’s GPU-kernel optimization run lasted seven days, explored more than 500 optimization directions, and produced 40 committed kernel versions.

2. 2

   AVO completed all 183 levels in the ARC-AGI-3 public set with a 100.00 RHAE score using 6,624 environment actions.

3. 3

   The ARC-AGI-3 comparison with VISTA is not a controlled ablation because the systems differ in backend, observations, memory, context management, and other implementation details.