AMAP Agentic Planning Technical Report

TL;DR

STAgent is a specialized large language model designed for complex spatio-temporal reasoning tasks, integrating multiple tools and a hierarchical training process to enhance performance while preserving general capabilities.

Contribution

The paper introduces a novel agentic LLM with a stable multi-tool environment, a hierarchical data curation framework, and a cascaded training recipe, advancing spatio-temporal reasoning capabilities.

Findings

Effective performance on TravelBench benchmark

Maintains general capabilities across various benchmarks

High-quality data curation improves training efficiency

Abstract

We present STAgent, an agentic large language model tailored for spatio-temporal understanding, designed to solve complex tasks such as constrained point-of-interest discovery and itinerary planning. STAgent is a specialized model capable of interacting with ten distinct tools within spatio-temporal scenarios, enabling it to explore, verify, and refine intermediate steps during complex reasoning. Notably, STAgent effectively preserves its general capabilities. We empower STAgent with these capabilities through three key contributions: (1) a stable tool environment that supports over ten domain-specific tools, enabling asynchronous rollout and training; (2) a hierarchical data curation framework that identifies high-quality data like a needle in a haystack, curating high-quality queries by retaining less than 1\% of the raw data, emphasizing both diversity and difficulty; and (3) a cascaded training recipe that starts with a seed SFT stage acting as a guardian to measure query difficulty, followed by a second SFT stage fine-tuned on queries with high certainty, and an ultimate RL stage that leverages data of low certainty. Initialized with Qwen3-30B-A3B to establish a strong SFT foundation and leverage insights into sample difficulty, STAgent yields promising performance on TravelBench while maintaining its general capabilities across a wide range of general benchmarks, thereby demonstrating the effectiveness of our proposed agentic model.