ProgAgent: Learning Tool Orchestration through Programmatic Training

Key points

• Large language models (LLMs) are increasingly expected to orchestrate diverse tools over long-horizon tasks.1
• Existing agents typically use Direct Tool Calling (DTC), where the model invokes tools sequentially, and tool responses accumulate in the model context, which can lead to context rot and orchestration breakdown.1
• Prior work has explored PTC mainly as an inference-time optimization, but its benefits for agentic training remain unclear, and suitable verifiable training tasks are scarce.1
• Cross-harness experiments show that PTC-trained models outperform DTC-trained models under both evaluation interfaces, suggesting that the benefits of PTC training extend beyond the interface.1

Sources (2)

• [1]ProgAgent: Learning Tool Orchestration through Programmatic Training
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 10, 09:57 AM

  In this work, we introduce ProgAgent, a full PTC training stack, comprising a pipeline for synthesizing verifiable long-horizon tasks, a lightweight PTC runtime, and a two-stage training recipe.

  Large language models (LLMs) are increasingly expected to orchestrate diverse tools over long-horizon tasks.

• [2]ProgAgent: Learning Tool Orchestration through Programmatic Training
  Hugging Face Daily Papers · Oct 10, 12:00 AM · same content

Extractive summary: sentences quoted from the sources.