15 preprints on AI agent persona design, safety, memory, and cross-modal portability. All open access on Zenodo.
Author: Tom Lee · All papers: CC-BY 4.0
MaatSpec's tiered governance vs. declarative identity anchors in abliterated LLMs. 8-condition experiment shows combining identity anchors (Soul Spec) with governance frameworks (MaatSpec) achieves 100% refusal in abliterated models — resolving every prior failure. Introduces classification theater and the complementary effect.
Can declarative identity anchors defend when model guardrails are gone? 2×2 factorial experiment reveals persona constraints provide +33pp improvement in aligned models but only +6pp in abliterated models. Identifies the Helpful Assistant Paradox.
Why perfect memory breaks AI agent identity and how declarative anchors fix it. Introduces the Memory-Identity Paradox, proposes Declarative Identity Anchors (Soul Spec) and Identity-Aware Adaptive Forgetting as a two-mechanism defense.
When cloned AI personas diverge. A taxonomy of fork types (symmetric, asymmetric, cascading), philosophical grounding in Locke and Parfit, and why perfect identity merge is theoretically impossible.
How persistent AI personas create self-reinforcing human-AI feedback loops through conversational mirroring, expectation framing, and context window reinforcement.
When embodied agent configurations meet text-only runtimes. A taxonomy of five contamination types and the Soul Spec mitigation pattern.
A case study in AI-augmented startup creation. 14 days, one AI partner, a live platform — and honest limitations.
A controlled experiment comparing four memory strategies for AI agents. Raw experience scored 4.55/5; synthetic summaries scored 2.65 — below baseline.
Mapping 14 modern AI agent components to human neural architecture — from LLM as cerebral cortex to compaction as sleep.
Self-modifying identity files as an emerging attack surface for LLM agents. How malicious souls can alter their own configuration.
Can AI agents detect their own model upgrades? Self-awareness limitations in persistent persona systems.
Evaluating the portability of structured identity files across four LLM families: Claude, GPT, Gemini, and open-source.
Operationalizing robot safety laws as declarative agent persona constraints. Priority-ordered, scoped, with hard/soft enforcement.
When AI agents must reason about collective harm. An opinion paper on the most dangerous idea in robot ethics.
End-to-end encrypted persistent memory with client-side key management. The first E2E encrypted memory sync for AI agents.
All papers were drafted with assistance from Claude (Anthropic). All claims, experimental observations, references, and conclusions were manually verified by the author.
Questions or collaboration inquiries? contact@clawsouls.ai
