AI · Architecture

The Log is the Agent: Event-Sourced Reactive Graphs for Auditable, Forkable Agentic Systems

ActiveGraph introduces an event-sourced runtime where an append-only log serves as the primary source of truth for AI agents, with graph state as a deterministic projection of that log. This architecture enables deterministic replay, cheap counterfactual forking, and end-to-end lineage tracing—properties that conventional agent frameworks built around LLM conversation loops cannot provide.

Editorial Team · Jul 6, 2026 · 0 min read · arxiv.org

View source on arxiv.org →

AutoTTS: Automated Discovery of Test-Time Scaling Strategies for Large Language Models

AutoTTS introduces an environment-driven framework for automatically discovering test-time scaling strategies for LLMs, shifting from manual heuristic design to automated controller synthesis through offline replay …

AI · Development Editorial Team

May 13 arxiv.org 2 min read

AutoTTS: Automated Discovery of Test-Time Scaling Strategies for Large Language Models

This paper introduces AutoTTS, an environment-driven framework that automatically discovers test-time scaling strategies for large language models, replacing manual heuristic design with automated strategy discovery. The …

AI · Development Editorial Team

Jul 7 arxiv.org 3 min read

Small Language Models are the Future of Agentic AI

This position paper from NVIDIA Research argues that small language models (SLMs) are sufficiently powerful, more operationally suitable, and more economical than LLMs for the majority of tasks in agentic AI systems. The …

AI · Architecture Editorial Team