Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
3 results for “long-horizon agents”
Toward Reliable Context Compression for Long-Horizon Agents: An Empirical Study of Execution Instability
A preliminary empirical study identifies instability risks in recurrent context compression for long-horizon AI agents and proposes TRACE, a verifier-guided framework that improves task performance and reliability without updating models.
Aug 10, 2026
SafeCommit: Certifying When Memory-Grounded Agents May Safely Act
SafeCommit is a new formal framework and risk-controlled layer designed to prevent AI agents from taking unsafe actions due to uncertain or flawed memory grounding by certifying commitment only when safety is guaranteed across a calibrated set of plausible latent worlds.
Aug 6, 2026
Oracle Agent Memory as an Enterprise Memory Substrate for Long-Horizon AI Agents
A new arXiv preprint introduces Oracle Agent Memory — a database-native memory substrate built on Oracle Database — designed to address long-horizon AI agent memory challenges through lifecycle management, layered architecture, and token-efficient evaluation.
Jul 16, 2026