POEM: Process Of Elimination Master — Introducing the Reductive Inference Model (RIM)
Abstract
We introduce the Reductive Inference Model (RIM), a new class of artificial intelligence architecture that reaches answers through progressive elimination of impossibilities rather than generative construction of possibilities. POEM (Process Of Elimination Master) is the first RIM. Unlike Large Language Models (LLMs), which activate full parameter sets at constant energy cost regardless of query difficulty, a RIM scales compute proportionally to actual query complexity. We describe the core architecture, its three primary components, and the training philosophy that enables this approach. The RIM paradigm requires no new hardware and represents a structural departure from the generative inference model that currently defines the field
Fields of Study & Subjects
Topic Modeling
Artificial Intelligence Applications
Natural Language Processing Techniques
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 5 March 2026
- Language
- EN
- Document type
- Preprint