Paper Companion
Accessible discussions of the formal research papers in the ARC-Eden programme. These posts explain what each paper proves, how it fits into the wider programme, and what evidence supports its claims. 23 posts
- ARC Principle Paper Suite: Master Table of Contents & Glossary | Michael Darius Eastwood This document provides a master table of contents, reading guide, timeline, key results summary, and glossary for the complete ARC Principle paper suite. It serves as the entry point for navigating the full research programme, listing all papers in recommended reading order with cross-references, an
- ARC-Align: a blind benchmark for depth-variable alignment A specification for measuring how AI alignment quality changes with reasoning depth, under blinded conditions and adversarial pressure.
- Capability Does Not Cancel Disability, PNP Framework Research | Michael Darius Eastwood The Polymathic Neurodivergent Profile (PNP): a falsifiable descriptive framework for the pattern of cross-domain high ability coexisting with disability-linked impairment. Capability does not cancel disability.
- Cauchy unification: an old mathematics as a scaling-law classifier A 200-year-old set of functional equations, tested as a cross-domain predictor of which scaling family fits which system.
- Eden Engineering: public research note Historical research note and withdrawal notice for the expanded Eden Engineering working specification.
- On the Origin of Scaling Laws - M. D. Eastwood (2026) | Michael Darius Eastwood A mouse's heart beats 600 times per minute; an elephant's beats 28; a blue whale's beats 6. This paper traces the origin of scaling laws across biological, physical, and computational systems, arguing that the recursive amplification mechanism identified by the ARC Principle provides a u
- Paper II: what the retraction actually says Paper II tried to pin down the ARC exponent experimentally. The first number was wrong. The revised number is smaller, harder to make headlines with, and more honest.
- Paper III: why external safety measures cannot keep up Paper III of the ARC/Eden programme argues that if capability compounds through recursion, external safety constraints that do not participate in that recursion get left behind.
- Paper IV-a: three tiers, not two, and one that reverses Paper IV-a of the ARC/Eden programme replaces the earlier baked-in/computed alignment binary with a three-tier hierarchy under blinded evaluation.
- Paper IV-b: alignment does not saturate the same way for every model Paper IV-b of the ARC/Eden programme reports that alignment saturation with reasoning depth is architecture-dependent, not universal.
- Synthesis and roadmap: an honest ledger of the ARC/Eden programme Eleven empirical studies, one positive result, two nulls, one inconclusive, twelve documented errors and a phased roadmap. Paper IX in plain English.
- The 18-paper suite: a reading order A plain-English map through the ARC/Eden research programme, honest about which results are strong, which are single-lab, and which are under recompute.
- The ARC Principle in plain English Paper I of Michael Darius Eastwood's ARC/Eden programme in plain English: U equals I times R to the alpha as a testable form for how test-time recursion scales capability. First published 17 January 2026 (OSF DOI 10.17605/OSF.IO/6C5XB), seeded in the 8 December 2024 manuscript.
- The convergence register: 22 claimed arrivals at the same principle Twenty-two claimed convergences (from a typed register of twenty-six dated events; four documented, not claimed) between the ARC/Eden framework and unrelated 2024 to 2026 research programmes. Paper XI in plain English.
- The ARC Co-Scaling Law: what actually keeps a self-improving AI safe Stability is set by an exponent inequality, correction must out-scale drift-acceleration, not by growth rate. Paper X in plain English, including what it supersedes.
- The Eden Protocol vision in plain English The philosophical companion to Eden Engineering: why a safe AI cannot be assembled by bolting rules onto a capable machine after the fact.
- The foundational ARC paper in plain English Three axioms plus Cauchy's 1821 theorem fix every recursive scaling law to one of three shapes. Which one you get is decided by the composition operator.
- The honey architecture: what happens when safety is baked into the objective A proposal that safety should be entangled with capability in the loss function itself, tested in toy self-modifying networks and pilot live-model batteries.
- The Hyperspace Recursive Intelligence Hypothesis - A Testable Creation Theory | Michael Darius Eastwood The Hyperspace Recursive Intelligence Hypothesis: a testable creation framing, held in the possibility register. Companion to the HRIH paper.
- The load-bearing test: what really survived the Eden Protocol tests One of three pre-specified experiments returned a positive result. Two returned null or inconclusive. Paper VIII in plain English.
- The programme executive summary in plain English A one-page tour of the ARC/Eden research programme for grant reviewers and general readers, with honest limits.
- The stewardship gene: caring about people as the foundational alignment trait A developmental-alignment argument: stakeholder care is the substrate trait from which nuance, honesty and quality can grow.
- When blinding flips the sign: why alignment scores need clinical-trial controls Four layers of blinding turned a +0.354 alignment improvement into a -0.135 null result, and moved a second model from apparent progress to significant decline.