Supplement Generation Training for Enhancing Agentic Task Performance
Young Min Cho, Daniele Bonadiman, Divya Bhargavi, Tamer Alkhouli, Salvatore Romeo + 6 more
TLDR
SGT trains small LLMs to generate supplemental text, boosting large LLM performance on agentic tasks without costly retraining.
Key contributions
- SGT trains a smaller LLM to create task-specific supplemental text.
- Appends generated supplements to input, enhancing large LLM task performance.
- Improves agentic task efficiency without modifying or retraining large foundation models.
- Decouples task optimization, enabling flexible and cost-effective LLM agent deployment.
Why it matters
This paper introduces SGT, a novel approach to efficiently enhance LLM agent performance. It addresses the high costs and rapid obsolescence of retraining large models by using lightweight supplements. This makes LLM-powered agents more adaptable and cost-effective for real-world applications.
Original Abstract
Training large foundation models for agentic tasks is increasingly impractical due to the high computational costs, long iteration cycles, and rapid obsolescence as new models are continuously released. Instead of post-training massive models for every new task or domain, we propose Supplement Generation Training (SGT), a more efficient and sustainable strategy. SGT trains a smaller LLM to generate useful supplemental text that, when appended to the original input, helps the larger LLM solve the task more effectively. These lightweight models can dynamically adapt supplements to task requirements, improving performance without modifying the underlying large models. This approach decouples task-specific optimization from large foundation models and enables more flexible, cost-effective deployment of LLM-powered agents in real-world applications.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.