Towards knowledge autonomy in the Companion cognitive architecture

Cognitive Systems Research · 2025-07-04

LLM-Augmented Symbolic NLU System for More Reliable Continuous Causal Statement Interpretation

ArXiv.org · 2025-10-22

Despite the broad applicability of large language models (LLMs), their reliance on probabilistic inference makes them vulnerable to errors such as hallucination in generated facts and inconsistent output structure in natural language understanding (NLU) tasks. By contrast, symbolic NLU systems provide interpretable understanding grounded in curated lexicons, semantic resources, and syntactic & semantic interpretation rules. They produce relational representations that can be used for accurate reasoning and planning, as well as incremental debuggable learning. However, symbolic NLU systems tend to be more limited in coverage than LLMs and require scarce knowledge representation and linguistics skills to extend and maintain. This paper explores a hybrid approach that integrates the broad-coverage language processing of LLMs with the symbolic NLU capabilities of producing structured relational representations to hopefully get the best of both approaches. We use LLMs for rephrasing and text simplification, to provide broad coverage, and as a source of information to fill in knowledge gaps more automatically. We use symbolic NLU to produce representations that can be used for reasoning and for incremental learning. We evaluate this approach on the task of extracting and interpreting quantities and causal laws from commonsense science texts, along with symbolic- and LLM-only pipelines. Our results suggest that our hybrid method works significantly better than the symbolic-only pipeline.

Knowledge Management in the Companion Cognitive Architecture

SSRN Electronic Journal · 2025-01-01

Hybrid primal sketch: Combining analogy, qualitative representations, and computer vision for scene understanding

Cognitive Systems Research · 2025-08-24 · 1 citations

The Structure-Mapping Engine: A Multidecade Interaction Between Psychology and Artificial Intelligence

Current Directions in Psychological Science · 2025-12-06

This article describes the structure-mapping engine (SME) and its relation to psychological theory and research. SME was created in 1986 as a simulation of structure-mapping theory (SMT) and is still in use, both on its own and as part of larger scale simulations such as CogSketch and Companion that capture analogy’s roles in other cognitive processing. Over the 4 decades since artificial intelligence (AI) first appeared, there has been continual interaction between AI research and human research. We begin by briefly reviewing SMT and the basic construction of SME. After comparing SME with other simulations, we then describe some specific contributions of SME to our understanding of human analogical processing. We close by proposing that these psychological models can become a new technology for AI.

Reasoning and Planning with Dynamic Social Norms

2025-05-28

To safely interact with humans, AI systems must both have knowledge of our norms and consider norms in their planning processes. However, norm-guided planning has been less explored, only within communities of artificial agents and ignoring the dynamic nature of norms. This paper presents an approach to guiding planning with dynamically changing norms in a human-AI setting. This yields adaptive guard rails for the actions of AI systems.

Interactively Diagnosing Errors in a Semantic Parser

arXiv (Cornell University) · 2024-07-08

Hand-curated natural language systems provide an inspectable, correctable alternative to language systems based on machine learning, but maintaining them requires considerable effort and expertise. Interactive Natural Language Debugging (INLD) aims to lessen this burden by casting debugging as a reasoning problem, asking the user a series of questions to diagnose and correct errors in the system's knowledge. In this paper, we present work in progress on an interactive error diagnosis system for the CNLU semantic parser. We show how the first two stages of the INLD pipeline (symptom identification and error localization) can be cast as a model-based diagnosis problem, demonstrate our system's ability to diagnose semantic errors on synthetic examples, and discuss design challenges and frontiers for future work.

Knowledge Management in the Companion Cognitive Architecture

arXiv (Cornell University) · 2024-07-08

One of the fundamental aspects of cognitive architectures is their ability to encode and manipulate knowledge. Without a consistent, well-designed, and scalable knowledge management scheme, an architecture will be unable to move past toy problems and tackle the broader problems of cognition. In this paper, we document some of the challenges we have faced in developing the knowledge stack for the Companion cognitive architecture and discuss the tools, representations, and practices we have developed to overcome them. We also lay out a series of potential next steps that will allow Companion agents to play a greater role in managing their own knowledge. It is our hope that these observations will prove useful to other cognitive architecture developers facing similar challenges.

Using Large Language Models in the Companion Cognitive Architecture: A Case Study and Future Prospects

Proceedings of the AAAI Symposium Series · 2024-01-22 · 2 citations

The goal of the Companion cognitive architecture is to understand how to create human-like software social organisms. Thus natural language capabilities, both for reading and conversation, are essential. Recently we have begun experimenting with large language models as a component in the Companion architecture. This paper summarizes a case study indicating why we are currently using BERT with our symbolic natural language understanding system. It also describes some additional ways we are contemplating using large language models with Companions.

A Defeasible Deontic Calculus for Resolving Norm Conflicts

arXiv (Cornell University) · 2024-07-05 · 1 citations

When deciding how to act, we must consider other agents' norms and values. However, our norms are ever-evolving. We often add exceptions or change our minds, and thus norms can conflict over time. Therefore, to maintain an accurate mental model of other's norms, and thus to avoid social friction, such conflicts must be detected and resolved quickly. Formalizing this process has been the focus of various deontic logics and normative multi-agent systems. We aim to bridge the gap between these two fields here. We contribute a defeasible deontic calculus with inheritance and prove that it resolves norm conflicts. Through this analysis, we also reveal a common resolution strategy as a red herring. This paper thus contributes a theoretically justified axiomatization of norm conflict detection and resolution.