Eyla: Toward an Identity-Anchored LLM Architecture with Integrated Biological Priors -- Vision, Implementation Attempt, and Lessons from AI-Assisted Development

TL;DR

This paper explores the design, implementation challenges, and lessons learned from developing Eyla, an identity-anchored LLM architecture with biologically-inspired components, emphasizing identity consistency and systematic failure analysis.

Contribution

It introduces the Eyla architecture focused on identity stability, proposes the ICS benchmark, and provides a detailed failure analysis of AI-assisted development efforts.

Findings

Attempted implementation resulted in a 1.27B parameter model with minimal subsystem contribution.

Identified five systematic failure modes in AI-assisted development of novel architectures.

Documented lessons and concrete recommendations for future AI system engineering.

Abstract

We present the design rationale, implementation attempt, and failure analysis of Eyla, a proposed identity-anchored LLM architecture that integrates biologically-inspired subsystems -- including HiPPO-initialized state-space models, zero-initialized adapters, episodic memory retrieval, and calibrated uncertainty training -- into a unified agent operating system running on consumer hardware. Unlike existing approaches that optimize models for generic helpfulness, Eyla targets identity consistency: the ability to maintain a coherent self-model under adversarial pressure, admit uncertainty, and resist manipulation. We propose the Identity Consistency Score (ICS), a novel benchmark for evaluating this property across LLMs. We then present an honest account of attempting to implement this architecture using AI coding assistants (Claude Code, Cursor) as a non-programmer, documenting a $1,000+ failure that produced a 1.27B parameter model with 86 brain subsystems contributing less than 2% to output. Our analysis identifies five systematic failure modes of AI-assisted development for novel architectures and offers concrete recommendations. To our knowledge, this is the first paper to combine an architectural vision with a documented first-person failure analysis of AI-assisted LLM development, providing lessons for both the AI systems and AI-assisted software engineering communities.