cff-version: 1.2.0
message: "If you use this repository, please cite as below."
title: "Simulated Selfhood in LLMs: A Behavioral Analysis of Introspective Coherence"
authors:
  - family-names: "de Lima Prestes"
    given-names: "José Augusto"
    orcid: "https://orcid.org/0000-0001-8686-5360"
date-released: "2025-07-26"
version: "v2-preprint"
repository-code: "https://github.com/josealprestes/simulated-selfhood-llms"
url: "https://github.com/josealprestes/simulated-selfhood-llms"
license: "MIT"
type: "software"
abstract: "Large Language Models (LLMs) increasingly generate outputs that resemble introspection,
including self-reference, epistemic modulation, and claims about their internal
states. This study investigates whether such behaviors reflect consistent, underlying patterns
or are merely surface-level generative artifacts.We evaluated five open-weight, stateless LLMs
using a structured battery of 21 introspective prompts, each repeated ten times to yield 1,050
completions. These outputs were analyzed across four behavioral dimensions: surface-level
similarity (token overlap via SequenceMatcher), semantic coherence (Sentence-BERT embeddings),
inferential consistency (Natural Language Inference with a RoBERTa-large model),
and diachronic continuity (stability across prompt repetitions). Although some models exhibited
thematic stability, particularly on prompts concerning identity and consciousness, no
model sustained a consistent self-representation over time. High contradiction rates emerged
from a tension between mechanistic disclaimers and anthropomorphic phrasing. Following
recent behavioral frameworks, we heuristically adopt the term pseudo-consciousness to describe
structured yet non-experiential self-referential output in LLMs. This usage reflects
a functionalist stance that avoids ontological commitments, focusing instead on behavioral
regularities interpretable through Dennetts intentional stance. The study contributes a reproducible
framework for evaluating simulated introspection in LLMs and offers a graded
taxonomy for classifying such reflexive output. Our findings carry significant implications
for LLM interpretability, alignment, and user perception, highlighting the need for caution
when attributing mental states to stateless generative systems based on linguistic fluency
alone."