Cognitive BASIC: An In-Model Interpreted Reasoning Language for LLMs

2025-11-20

AI Summary

This work addresses the opacity and traceability challenges inherent in large language model (LLM) inference. We propose Cognitive BASIC—a structured, BASIC-inspired reasoning language that introduces a human-interpretable cognitive control layer within LLMs. It employs numbered lines, atomic instructions, and explicit memory operations to enable stepwise, executable thought tracing. Uniquely integrating retro-style procedural programming with natural-language semantic definitions, Cognitive BASIC supports in-prompt explanation-aware execution, dynamic logging, and joint declarative-procedural knowledge extraction. Experimental evaluation across three mainstream LLMs demonstrates substantial improvements in inference transparency and robust performance gains on knowledge extraction, logical conflict detection, and multi-step reasoning tasks. Our approach establishes a novel paradigm for controllable, auditable, and interpretable LLM reasoning—bridging symbolic clarity with neural expressivity while preserving end-to-end differentiability.

Technology Category

• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Natural Language Processing: (Large) Language Models
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Cognitive BASIC is a minimal, BASIC-style prompting language and in-model interpreter that structures large language model (LLM) reasoning into explicit, stepwise execution traces. Inspired by the simplicity of retro BASIC, we repurpose numbered lines and simple commands as an interpretable cognitive control layer. Modern LLMs can reliably simulate such short programs, enabling transparent multi-step reasoning inside the model. A natural-language interpreter file specifies command semantics, memory updates, and logging behavior. Our mental-model interpreter extracts declarative and procedural knowledge, detects contradictions, and produces resolutions when necessary. A comparison across three LLMs on a benchmark of knowledge extraction, conflict detection, and reasoning tasks shows that all models can execute Cognitive BASIC programs, with overall strong but not uniform performance.

Problem

• Structures LLM reasoning into explicit stepwise execution traces
• Extracts declarative and procedural knowledge while detecting contradictions
• Enables transparent multi-step reasoning through BASIC-style interpreted language

Innovation

• Minimal BASIC-style language structures LLM reasoning
• In-model interpreter enables transparent stepwise execution
• Natural-language specification defines command semantics and memory