Overview
The AlphaFund whitepaper reframes recursive self-improvement (RSI) as a portfolio
optimization problem: a corporation recursively improves when realized economic
gains finance the next cycle of better prediction and deployment, and the firm's
standing is summarized by t-RSI, a standardized gap between alpha-creation and
alpha-decay rates. AlphaCOGANT observes that this construction is, term for
term, an Active Inference agent — and makes the correspondence executable.
We render AlphaFund's Economic World Model (EWM) as a generative model written
in Generalized Notation Notation (GNN), produced by the COGANT
codebase-to-GNN translation pattern. The firm's five capital channels —
Investments, Sensors, Actuators, Parameters, and R&D — become the hidden-state
factors of a partially-observed model; capital allocation becomes the control
vector; and the portfolio optimizer's marginal-return objective becomes
Expected Free Energy (EFE) minimization. The EFE decomposition supplies a
principled reading of AlphaFund's own categories: its pragmatic value is
expected log-equity growth (the alpha-creation rate, read off the broker ledger),
and its epistemic value is the information gain about the EWM that Sensors and
R&D purchase (the data-scaling and forecast-sharpening laws). t-RSI is recovered
as the standardized distance between the create-rate and decay-rate posteriors —
the thresholded EFE-improvement certificate that admits a self-improvement commit
only when creation confidently exceeds decay.
We give the technical and computational realization: a GNN model file for the
five-channel firm, a tested NumPy Active Inference engine that performs state
inference, computes the epistemic/pragmatic EFE split and the marginal-return
vector, and evaluates the t-RSI certificate. We argue that GNN-via-COGANT brings
two things AlphaFund's program needs and Active Inference already enforces:
filtration integrity (the model may condition only on information available
at decision time — the same "no-peeking" discipline that separates an EWM from a
language model) and auditable capital allocation (every admissible funding
move has a negative-EFE score under a single, legible objective). This is not
financial advice; it is a demonstration that this reduced
recursive-corporate-self-improvement model has a direct Active Inference
representation supported by source-owning methods and artifact checks .
---
Associated artifacts
GitHub release: v1.0.1 (https://github.com/docxology/alphacogant/releases/tag/v1.0.1)
DOI: https://doi.org/10.5281/zenodo.20976824
Zenodo: https://zenodo.org/records/20976824
PDF SHA-256: 41efa7a8a98e6a67cece68377b8f0c5f19304e85c664ec9516353ee24eb0421f
Overview source: Curated paper metadata.
Methods and contributions
Read the source for the full argument, qualifications, and evidence.
Findings and contributions
- The paper argues AlphaFund's recursive-self-improvement-as-portfolio-optimization has an Active Inference representation that is expressible in GNN and producible by the COGANT pattern.
- t-RSI is recovered as the standardized distance between create-rate and decay-rate posteriors, i.e. a thresholded EFE-improvement certificate gating self-improvement commits.
- At point-estimate level the comparator discriminates: create exceeds decay (admit) at the self-improving point and falls below it (reject) at the coasting point.
- Caveat: with bootstrapped uncertainty the reduced two-level model's headline t-RSI is negative (-13.2552) at the self-improving point, so it does not robustly certify net improvement.
- The authors state AlphaCOGANT is a modeling and integrity instrument, not a trading system or financial advice, and does not reproduce AlphaFund's proprietary surfaces or track record.
Methods
- Construct-by-construct AlphaFund-to-Active Inference dictionary — Mapped each AlphaFund whitepaper construct (corporation tuple, EWM, filtration, action vector, t-RSI, etc.) to an Active Inference object.
- GNN model file of the five-channel firm produced via the COGANT pattern — Wrote AlphaFund's Economic World Model as a Generalized Notation Notation file with channel factors, A/B matrices, log-preferences and an EFE objective, using COGANT's codebase-to-GNN step.
- Deterministic tested NumPy Active Inference engine (src/alphacogant/) — Implemented state inference over channels, the epistemic/pragmatic EFE split, the marginal-return vector, and t-RSI certificate evaluation.
- Engine-gated manuscript numbers and figure-provenance registry — Generated every cited numeric from one function and cross-checked it by test, and registered each figure with its producer script and hashed manifest.
- Bootstrap of create/decay posteriors with Dirichlet concentration sensitivity — Bootstrapped belief uncertainty to compute standardized t-RSI and reported its sensitivity to the Dirichlet concentration parameter alpha.
Summary sources: Paper metadata and evidence · Extracted source text.
Document downloads
PDFs are hosted here; supporting sources may use pinned archives.
- Friedman_2026_Alphacogant_41efa7a8.pdf PDF · 2.75 MiB · extracted-text source
Citation
Citation metadata follows the unified bibliography.
Catalog details and resources
Related works
Ranked by recorded methods, topics, contributors and associated software, with contributor roles and domain-only matches labelled. Metadata overlap is not a citation or a scientific equivalence.