T1: A Tool-Oriented Conversational Dataset for Multi-Turn Agentic Planning

TL;DR

T1 is a comprehensive multi-domain conversational dataset designed to evaluate and improve large language models' ability to plan, coordinate, and manage tool dependencies in multi-turn, multi-agent scenarios.

Contribution

The paper introduces T1, a novel dataset for multi-turn, multi-domain agentic planning that includes mechanisms for caching and dynamic replanning, advancing research in tool use by LLMs.

Findings

T1 enables effective evaluation of multi-tool coordination.

T1-Agent demonstrates strong planning and reasoning capabilities.

The dataset supports benchmarking of various LLMs in complex scenarios.

Abstract

Large Language Models (LLMs) have demonstrated impressive capabilities as intelligent agents capable of solving complex problems. However, effective planning in scenarios involving dependencies between API or tool calls-particularly in multi-turn conversations-remains a significant challenge. To address this, we introduce T1, a tool-augmented, multi-domain, multi-turn conversational dataset specifically designed to capture and manage inter-tool dependencies across diverse domains. T1 enables rigorous evaluation of agents' ability to coordinate tool use across nine distinct domains (4 single domain and 5 multi-domain) with the help of an integrated caching mechanism for both short- and long-term memory, while supporting dynamic replanning-such as deciding whether to recompute or reuse cached results. Beyond facilitating research on tool use and planning, T1 also serves as a benchmark for evaluating the performance of open-weight and proprietary large language models. We present results powered by T1-Agent, highlighting their ability to plan and reason in complex, tool-dependent scenarios.