Graph-Augmented Episodic Memory for Constraint-Aware Agentic Planning
Integrate a graph-based episodic memory and Monte Carlo Tree Search into the agent planning loop to handle long-horizon sequential state dependencies.
Source paper
Shiva Krishna Reddy Malay, Shravan Nayak, Jishnu Sethumadhavan Nair, Sagar Davasam, Aman Tiwari, Sathwik Tejaswi Madhusudhan, Sridhar Krishna Nemala, Srinivas Sunkara, Sai Rajeswar
Generated by AI. The paper’s authors did not write these.
Integrate a graph-based episodic memory and Monte Carlo Tree Search into the agent planning loop to handle long-horizon sequential state dependencies.
Train a lightweight process reward model to score policy compliance and trigger safe refusals when agents encounter infeasible tasks.
Implement a domain-grounded test-time compute router that injects procedural RAG rules into the reasoning phase to satisfy strict constraint-heavy domains.
This paper introduces ENTERPRISEOPS-GYM, a highly complex, sandboxed benchmark featuring 164 database tables and 512 tools designed to evaluate LLM agents in realistic, policy-constrained enterprise workflows. The study reveals that current frontier models struggle significantly with long-horizon strategic planning, cascading state dependencies, and safely refusing infeasible tasks, indicating they are not yet ready for autonomous enterprise deployment.
The authors developed a fully containerized simulation covering eight enterprise domains (e.g., HR, ITSM, CSM). They evaluated 14 frontier models using a ReAct-style reasoning and tool-execution loop in an oracle-tool setting, measuring success via expert-authored, outcome-based SQL verifiers that enforce goal completion, state integrity, and strict policy compliance.