Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Foundation Model Sherpas: Guiding Foundation Models through Knowledge and Reasoning

Debarun Bhattacharjya, Junkyu Lee, Don Joven Agravante, Balaji Ganesan, Radu Marinescu

TL;DR

Foundation models exhibit strong capabilities but pose trust, knowledge integration, and robustness challenges for real-world use. The authors propose the sherpas framework, a taxonomy of agent roles (FM Updaters, Prompt Assistants, Assessors, Knowledge Curators, Orchestrators) and four interaction protocols to steer FMs via knowledge augmentation and reasoning. By mapping contemporary approaches to these roles, the work clarifies how multi-agent collaboration can enhance trustworthiness, adaptability, and practicality in deployed systems. The framework lays out autonomous, proactive guiding agents and benchmarks to evaluate system-level capabilities beyond task-specific performance.

Abstract

Foundation models (FMs) such as large language models have revolutionized the field of AI by showing remarkable performance in various tasks. However, they exhibit numerous limitations that prevent their broader adoption in many real-world systems, which often require a higher bar for trustworthiness and usability. Since FMs are trained using loss functions aimed at reconstructing the training corpus in a self-supervised manner, there is no guarantee that the model's output aligns with users' preferences for a specific task at hand. In this survey paper, we propose a conceptual framework that encapsulates different modes by which agents could interact with FMs and guide them suitably for a set of tasks, particularly through knowledge augmentation and reasoning. Our framework elucidates agent role categories such as updating the underlying FM, assisting with prompting the FM, and evaluating the FM output. We also categorize several state-of-the-art approaches into agent interaction protocols, highlighting the nature and extent of involvement of the various agent roles. The proposed framework provides guidance for future directions to further realize the power of FMs in practical AI systems.

Foundation Model Sherpas: Guiding Foundation Models through Knowledge and Reasoning

TL;DR

Foundation models exhibit strong capabilities but pose trust, knowledge integration, and robustness challenges for real-world use. The authors propose the sherpas framework, a taxonomy of agent roles (FM Updaters, Prompt Assistants, Assessors, Knowledge Curators, Orchestrators) and four interaction protocols to steer FMs via knowledge augmentation and reasoning. By mapping contemporary approaches to these roles, the work clarifies how multi-agent collaboration can enhance trustworthiness, adaptability, and practicality in deployed systems. The framework lays out autonomous, proactive guiding agents and benchmarks to evaluate system-level capabilities beyond task-specific performance.

Abstract

Foundation models (FMs) such as large language models have revolutionized the field of AI by showing remarkable performance in various tasks. However, they exhibit numerous limitations that prevent their broader adoption in many real-world systems, which often require a higher bar for trustworthiness and usability. Since FMs are trained using loss functions aimed at reconstructing the training corpus in a self-supervised manner, there is no guarantee that the model's output aligns with users' preferences for a specific task at hand. In this survey paper, we propose a conceptual framework that encapsulates different modes by which agents could interact with FMs and guide them suitably for a set of tasks, particularly through knowledge augmentation and reasoning. Our framework elucidates agent role categories such as updating the underlying FM, assisting with prompting the FM, and evaluating the FM output. We also categorize several state-of-the-art approaches into agent interaction protocols, highlighting the nature and extent of involvement of the various agent roles. The proposed framework provides guidance for future directions to further realize the power of FMs in practical AI systems.
Paper Structure (26 sections, 2 figures, 1 table)

This paper contains 26 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Agents and Foundation Models
  3. Agent Roles in the Sherpas Framework
  4. FM Updaters
  5. Prompt Assistants and Sequencers
  6. Instruction-based Prompts
  7. In-context Learning Prompts
  8. Decomposed Prompts
  9. Assessors and Explainers
  10. Harmfulness/Toxicity.
  11. Factuality/Faithfulness.
  12. Consistency.
  13. Uncertainty/Confidence.
  14. Hallucination.
  15. Bias.
  16. ...and 11 more sections

Figures (2)

  • Figure 1: The sherpas framework for guiding FMs, showing various agent categories with respect to their typical points of interaction with the FM as it executes or assists completion of a set of tasks.
  • Figure 2: A relatively high-level taxonomy of four major sherpa categories, along with some illustrative example references.