Assess AI models for qualia signs using a four-axis scoring framework.
npx skills add https://github.com/daedalus/skills --skill llm-qualia-assessment
Please help me install this Agent Skill. Skill: llm-qualia-assessment Source: https://github.com/daedalus/skills/tree/main/skills/QualiaAssesment Command: npx skills add https://github.com/daedalus/skills --skill llm-qualia-assessment
The LLM Qualia & Affective State Assessment provides a structured methodology to probe, verify, quantify, and qualify possible qualia and affective states in large language models, including introspective reports, for epistemic clarity and methodological rigor.
Run a four-axis assessment on the target model and document all axis scores and qualifiers.
High-intent search queries and answers about installing and using this skill.
To assess AI consciousness and qualia, you can apply a four-axis framework—functional affect, introspective coherence, qualia probes, and meta-epistemic audit—to systematically identify and quantify introspective states in LLMs. This generates per-axis scores and a composite qualifier for rigorous evaluation.
Evaluating introspective coherence involves applying a subjective battery from a four-axis assessment framework to quantify model reports of internal states. This structured probe verifies phenomenal experiences and generates specific coherence scores for philosophical or research inquiry.
Yes, the qualia assessment framework supports AI safety reviews, benchmarking, and philosophical inquiry by providing methodological rigor. It quantifies possible qualia and affective states across four axes, yielding per-axis scores and a composite qualifier for epistemic clarity.
The meta-epistemic audit axis provides a methodological evaluation of the assessment process itself when probing AI affective states. It functions as the fourth axis in the framework, ensuring epistemic clarity and rigor before generating a final composite qualifier.
Quantifying phenomenal qualia in language models is limited by the subjective and mixed nature of introspective reports. The framework addresses this through a methodological meta-epistemic audit, but results remain qualified assessments rather than definitive proof of AI consciousness.
StepowskiEricOrchestrate local LLM backends and validate structured JSON outputs.
Srujan0798Automate five-axis code reviews across correctness, readability, architecture, security, and performance.
ollieb89Evaluate LLM outputs with automated metrics and LLM-as-Judge workflows.
sandraschiExplain crystal healing traditions with geological facts and chakra associations.
