docs: canonical acronym expansions

DESIGN.md

@@ -1,6 +1,6 @@
 # MADDENING v1 Design Document
 
-**Modular Automatic Differentiation and Data-Enhanced Neural-network INteracting Graph**
+**Modular Automatic Differentiation and Data Enhanced Neural-network INteracting Graph**
 
 ## Overview
 

DOCUMENTATION_ARCHITECTURE.md

@@ -2,10 +2,10 @@
 
 ## Executive Summary
 
-**MADDENING** (Modular Automatic Differentiation and Data-Enhanced Neural-network INteracting Graph) is a JAX-based HPC framework for multiphysics simulation. It occupies a dual-purpose position that demands careful documentation architecture from the start:
+**MADDENING** (Modular Automatic Differentiation and Data Enhanced Neural-network INteracting Graph) is a JAX-based HPC framework for multiphysics simulation. It occupies a dual-purpose position that demands careful documentation architecture from the start:
 
 1. **Research tool** — used directly by scientists and engineers to build, train, and run multiphysics simulations with automatic differentiation and neural surrogates.
-2. **Auditable foundation** — the computational backbone of **MIME** (MIcrorobotics Multiphysics Engine), which in turn powers **MICROBOTICA** (MICROROBOTs Iterative Simulation for Clinical Adoption), an open-source research simulator for microrobot-assisted drug delivery in cerebrospinal fluid and other confined biological geometries. MADDENING, MIME, and MICROBOTICA are all open-source research tools. The regulated clinical product — the thing that actually gets CE-marked and used in a clinical setting — will be built by a downstream commercial entity (spin-out, licensee, or partner company) on top of these open-source tools.
+2. **Auditable foundation** — the computational backbone of **MIME** (MIcrorobotics Multiphysics Engine), which in turn powers **MICROBOTICA** (MICROROBOTics Iterative simulation for Clinical Adoption), an open-source research simulator for microrobot-assisted drug delivery in cerebrospinal fluid and other confined biological geometries. MADDENING, MIME, and MICROBOTICA are all open-source research tools. The regulated clinical product — the thing that actually gets CE-marked and used in a clinical setting — will be built by a downstream commercial entity (spin-out, licensee, or partner company) on top of these open-source tools.
 
 This dual purpose means MADDENING's documentation must serve two audiences simultaneously: researchers who need to understand, extend, and publish with the framework, and downstream commercial manufacturers who need to cite MADDENING's verification record in regulatory submissions — without MADDENING itself making any clinical claims.
 
@@ -770,7 +770,7 @@ Test requirements at three levels:
 ```markdown
 # MADDENING
 
-Modular Automatic Differentiation and Data-Enhanced Neural-network
+Modular Automatic Differentiation and Data Enhanced Neural-network
 INteracting Graph.
 
 [Badges: CI status, test count, coverage, PyPI version, license]
@@ -2045,7 +2045,7 @@ MADDENING should publish a `docs/validation/soup_package.md` with each release,
 | Field | Value |
 |-------|-------|
 | Name | MADDENING |
-| Full name | Modular Automatic Differentiation and Data-Enhanced Neural-network INteracting Graph |
+| Full name | Modular Automatic Differentiation and Data Enhanced Neural-network INteracting Graph |
 | Version | [X.Y.Z] |
 | Release date | [YYYY-MM-DD] |
 | License | LGPL-3.0-or-later |

docs/index.md

@@ -1,8 +1,8 @@
 # MADDENING
 
-**Modular Acausal Dataflow Differential Equation Node Network** — a pure-JAX,
-autodifferentiable framework for composing physical models from typed nodes
-connected by unit-aware edges. MADDENING is the base framework on which
+**Modular Automatic Differentiation and Data Enhanced Neural-network INteracting
+Graph** — a pure-JAX, autodifferentiable framework for composing physical models
+from typed nodes connected by unit-aware edges. MADDENING is the base framework on which
 [MIME](https://microrobotica.org/mime/) and the
 [MICROROBOTICA](https://microrobotica.org/) IDE are built.
 

docs/regulatory/intended_use.md

@@ -2,7 +2,7 @@
 
 ## Platform Positioning Statement
 
-MADDENING (Modular Automatic Differentiation and Data-Enhanced Neural-network INteracting Graph) is a **general-purpose computational framework** for multi-physics simulation. It provides infrastructure for composing, coupling, and differentiating physics simulations using JAX, with support for neural surrogates and uncertainty quantification.
+MADDENING (Modular Automatic Differentiation and Data Enhanced Neural-network INteracting Graph) is a **general-purpose computational framework** for multi-physics simulation. It provides infrastructure for composing, coupling, and differentiating physics simulations using JAX, with support for neural surrogates and uncertainty quantification.
 
 **MADDENING is not a medical device** as defined by EU MDR (EU 2017/745) Article 2(1). It does not have a medical purpose, does not make clinical predictions, and does not provide diagnostic, therapeutic, or monitoring functionality. Per MDCG 2019-11 Section 3.2, MADDENING is a general-purpose computing tool comparable to MATLAB, NumPy, or FEniCS — it can be used to build medical device software, but is not itself a medical device.
 

docs/validation/soup_package.md

@@ -8,7 +8,7 @@
 | Field | Value |
 |-------|-------|
 | Name | MADDENING |
-| Full Name | Modular Automatic Differentiation and Data-Enhanced Neural-network INteracting Graph |
+| Full Name | Modular Automatic Differentiation and Data Enhanced Neural-network INteracting Graph |
 | Version | 0.1.0 |
 | Release Date | 2025-03-01 |
 | Licence | LGPL-3.0-or-later |

src/maddening/__init__.py

@@ -1,5 +1,5 @@
 """
-MADDENING - Modular Automatic Differentiation and Data-Enhanced
+MADDENING - Modular Automatic Differentiation and Data Enhanced
              Neural-network INteracting Graph
 
 A JAX-based modular simulation framework for multi-physics.