Multi-Turn Human-LLM Interaction Through the Lens of a Two-Way Intelligibility Protocol

TL;DR

This paper introduces a structured two-way intelligibility protocol for human-LLM interactions, demonstrating its effectiveness in complex data analysis tasks like radiology and drug design through empirical experiments.

Contribution

It presents a novel protocol based on finite-state machines for structured human-LLM interaction, with implementation and empirical validation in scientific domains.

Findings

Protocol effectively captures one- and two-way intelligibility.

Empirical evidence shows improved interaction quality.

Code implementation is publicly available.

Abstract

Our interest is in the design of software systems involving a human-expert interacting -- using natural language -- with a large language model (LLM) on data analysis tasks. For complex problems, it is possible that LLMs can harness human expertise and creativity to find solutions that were otherwise elusive. On one level, this interaction takes place through multiple turns of prompts from the human and responses from the LLM. Here we investigate a more structured approach based on an abstract protocol described in [3] for interaction between agents. The protocol is motivated by a notion of "two-way intelligibility" and is modelled by a pair of communicating finite-state machines. We provide an implementation of the protocol, and provide empirical evidence of using the implementation to mediate interactions between an LLM and a human-agent in two areas of scientific interest (radiology and drug design). We conduct controlled experiments with a human proxy (a database), and uncontrolled experiments with human subjects. The results provide evidence in support of the protocol's capability of capturing one- and two-way intelligibility in human-LLM interaction; and for the utility of two-way intelligibility in the design of human-machine systems. Our code is available at https://github.com/karannb/interact.