Deployed, autonomous AI systems must often evaluate multiple plausible courses of action (extended sequences of behavior) in novel or under-specified contexts. Despite extensive training, these systems will inevitably encounter scenarios where no available course of action fully satisfies all operational constraints (e.g., operating procedures, rules, laws, norms, and goals). To achieve goals in accordance with human expectations and values, agents must go beyond their trained policies and instead construct, evaluate, and justify candidate courses of action. These processes require contextual ``knowledge'' that may lie outside prior (policy) training. This paper characterizes requirements for agent decision making in these contexts. It also identifies the types of knowledge agents require to make decisions robust to agent goals and aligned with human expectations. Drawing on both analysis and empirical case studies, we examine how agents need to integrate normative, pragmatic, and situational understanding to select and then to pursue more aligned courses of action in complex, real-world environments.

Large language models (LLMs) offer significant promise as a knowledge source for task learning. Prompt engineering has been shown to be effective for eliciting knowledge from an LLM, but alone it is insufficient for acquiring relevant, situationally grounded knowledge for an embodied agent learning novel tasks. We describe a cognitive-agent approach, STARS, that extends and complements prompt engineering, mitigating its limitations and thus enabling an agent to acquire new task knowledge matched to its native language capabilities, embodiment, environment, and user preferences. The STARS approach is to increase the response space of LLMs and deploy general strategies, embedded within the autonomous agent, to evaluate, repair, and select among candidate responses produced by the LLM. We describe the approach and experiments that show how an agent, by retrieving and evaluating a breadth of responses from the LLM, can achieve 77-94% task completion in one-shot learning without user oversight. The approach achieves 100% task completion when human oversight (such as an indication of preference) is provided. Further, the type of oversight largely shifts from explicit, natural language instruction to simple confirmation/discomfirmation of high-quality responses that have been vetted by the agent before presentation to a user.

Patterns of life (POL) are emergent properties of com- plex social systems such as neighborhoods or even cities. Accurately modeling POL is not only an academic pursuit; complex training and analysis efforts rely upon these models. Computational POL models will benefit academic researchers by giving them new tools to generate and validate theories about complex sociocultural patterns. POL presents an interesting computational modeling challenge for research in many AI fields because of its complex, multi-level interaction. POL modeling differs from standard agent-based modeling in that not only individual behaviors, but also their collective emergent patterns, are targets of study. At the same time, the behavior of any single individual is potentially important.

Hierarchical task decomposition is a method used in many agent systems to organize agent knowledge. This work shows how the combination of a hierarchy and persistent assertions of knowledge can lead to difficulty in maintaining logical consistency in asserted knowledge. We explore the problematic consequences of persistent assumptions in the reasoning process and introduce novel potential solutions. Having implemented one of the possible solutions, Dynamic Hierarchical Justification, its effectiveness is demonstrated with an empirical analysis.

In this paper we discuss the nature of our overall enterprise to create ontologies in the product and service knowledge space for Business-to-Business (B2B) electronic commerce. We describe one crucial problem: the mapping problem, i.e., mapping among ontologies, taxonomies, and classification systems, some of which are more semantically sound and coherent than others. This problem we consider to be in need of a sustained research program if tenable solutions are to be found, since the lack of a solution will preclude widespread adoption of ontologies by the commercial world. Finally, we summarize the general issues we faced and indicate prospective future research.