🤖 AI Summary
Current large language model (LLM) workflow systems lack semantic representation and persistence mechanisms for workflows themselves, hindering inspectability, recoverability, and auditability. This work proposes a language-agnostic conceptual model inspired by Lisp that treats workflows as knowledge objects rather than mere execution traces. By leveraging symbolic forms, object identity, and the notion of live mirrors, the model distinguishes deterministic computation (derive) from LLM-mediated judgment (infer). It further integrates contextual snapshots and capability policies to govern reasoning processes. This approach establishes a semantic persistence framework for LLM workflows, laying preliminary formal foundations and substantially enhancing their inspectability, recoverability, and auditability.
📝 Abstract
Large language model (LLM) applications increasingly use explicit workflows for tool use, retrieval, branching, checkpointing, and human approval. Existing workflow systems already address many execution concerns. This paper proposes a Lisp-inspired but language-independent conceptual model: symbolic forms, object identity, and live-image thinking are used as explanatory lenses, not implementation commitments. In this model, workflow definitions, workflow instances, inference records, context snapshots, and dependency relations are represented as persistent knowledge objects in a shared knowledge substrate. Its central semantic distinction is between derive and infer: derive is deterministic computation over available state; infer is mediated LLM judgment under declared context and executor-controlled capability policy.
The result is a preliminary conceptual account of semantic persistence: workflows do not merely produce knowledge and leave traces, but can themselves be represented as inspectable, resumable, and reviewable knowledge objects, while formal transition semantics remain future work.