CORTEXA
← Browse
arxivcs.NE2026-07-18

How to Build Marcus's Algebraic Mind: From Minsky's Emotion-Machine Viewpoint

Hiroyuki Chuma, Kanji Otsuka, Yoichi Sato

In The Algebraic Mind, Marcus identified three cognitive components: operations over variables, recursively structured representations, and an individual/kind distinction. He left the neural substrate open. A companion paper solves this with VaCoAl, an architecture built on GF(2) XOR-and-shift. It offers exact reversible variable binding, compositional bundling, and spatial individual/kind separation. A collision-avoidance scheme ensures exact recovery within a Frontier Size. Its PyVaCoAl implementation demonstrates this algebra scales to millions of records, while the SRAM-CAM version favors hardware speed. Marcus's horizontal framework concerns how a mind represents the world. This paper reads VaCoAl through an orthogonal lens: Minsky's The Emotion Machine, providing the vertical dimension of a mind reasoning about itself. We claim VaCoAl's exact preservation and recovery of structural representations directly enables introspection -- Minsky's Reflective layer. Reflexively unbinding one's own deliberative trace is the operation. We grade deliberation into three tiers by trace faithfulness: probabilistic (LLMs), approximate-algebraic, and exact-algebraic (VaCoAl). Only exact reversibility ensures reflexive recovery is faithful rather than spurious, reaching the Reflective layer. This substrate carries two more Minsky ideas: panalogy (achieved via content-addressable retrieval where match accuracy indicates analogical fit, enabling lifelong learning) and credit assignment (framed as reflexive counterfactual simulation: surgically unbinding a past choice and rebinding a counterfactual to compare). The exact reversibility allowing the companion paper to climb Pearl's causal ladder helps this paper ascend Minsky's introspective stack, finding the intersection. We separate demonstrated trace execution from theoretical Reflection arguments, leaving meta-control loops strictly for future work.

View free PDFSource page

Related papers

arxivcs.NE2026-07-23

Search Hardness-Aware LLM-Based Problem Formulation for Expensive Simulation-Driven Design

Yuchen Li, Handing Wang, Bing Xue, Mengjie Zhang

Expensive simulation-driven design is widely used in engineering to identify requirement-satisfying designs with as few high-fidelity simulations as possible. Most existing efforts address this challenge by improving optimization algorithms under fixed formulations, yet the formu…

View free PDFSource page
arxivcs.LGcs.NE2026-07-23

Weight-norm Criticality: A Mechanism for Loss Spikes Induced by the Normalization and Weight Decay

Xiaolong Li, Zhangchen Zhou, Zhi-Qin John Xu

Most explanations of training instability focus on \emph{learning-rate criticality}, typically characterized by the Edge of Stability, beyond which optimization becomes unstable. We argue that, in practical deep neural network training, there is an additional and often overlooked…

View free PDFSource page