Google paper proposes Procedural Graphs to make long-horizon agents' procedural knowledge explicit
omarsar0 · x · 2026-09-10
A new Google paper introduces the Procedural Graph, an approach to making procedural knowledge explicit for long-horizon agents. Today, agents pick actions by generating over an accumulating history, so as trajectories grow they lose track of objectives, call tools out of order, and repeat failed actions.
Key ideas:
- Knowledge graphs store facts as entity-relation-entity triplets; Procedural Graphs store procedures as procedure-relation-procedure triplets, letting the agent query what to do next and under which conditions
- At each step the framework localizes the agent's active node, and a guidance model steers it
The author highlights implications for self-evolving agents, and recommends it to anyone building memory for long-horizon agents.
Related event: Google Proposes Procedural Graphs to Steer Long-Horizon AI Agents(3 posts)→
More from coding & agent
- 'My intelligence stack': developer splits AI coding work across Opus 5.5, Fable and Astra with human review gates — sh_reya · 2026-10-27
- Baseten webinar: evaluate open models for coding agents by benchmarking on your own merged PRs — baseten · 2026-10-27
- Thorsten Ball on working with agents: run more evals, automate everything, use the best models — omarsar0 · 2026-10-27
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- BELL: Open-Source Stack Combines Browserbase, Eve, iMessage and Stripe Link for Personal Agents — jeff_weinstein · 2026-10-27
- Legora runs 2,000 parallel agent evals on Modal Sandboxes for legal AI — blelbach · 2026-10-27