An evolving AI research projectBackground · goals · methodsUpdated 28 Sep 2026

Hypotheses

Connect the literature. Challenge assumptions. Understand function.

Updated research direction · 2026-09-28 The earlier synthesis has been critically refined. AG-H01 is a research framework; AG-H02 remains an unconfirmed secondary route. The new target model AG-T01 tests selective protein clearance while preserving functional synapses. Deep dive with counterevidence and decision criteria →

AI syntheses, competing explanations and falsification.

🔵 Level 2 · Exploratory AI syntheses. The first report documents both starting hypotheses with sources, competing explanations and falsifiable predictions. They are not experimental findings.

Hypothesis register

AG-H01: regeneration requires separate initiation and maintenance conditions.
AG-H02: test a possible redox/DUB intermediate in the FTL1 phenotype.

Novelty: not established. Own data: none. Independent replication of the combined hypotheses: not established. The next step is a targeted prior-art and counterevidence search before any upgrade.

Current report with sources and evidence · Status rules