← All programmes Programme V

Computational Cognition

Machine learning, safety, phenomenology, and the computational basis of cognitive processes. Trauma as training data corruption. Genre mimicry in alignment.

4 papers · hosted on dissensus.ai
Preprint 2025

Genre Mimicry vs. Ethical Reasoning in Abliterated Language Models

Murad Farzulla

Zenodo 10.5281/zenodo.17957693
Read on dissensus.ai →
Preprint 2025

Autonomous Red Team AI: LLM-Guided Adversarial Security Testing

Murad Farzulla, Andrew Maksakov

Zenodo 10.5281/zenodo.17614725
Read on dissensus.ai →
Preprint 2025

Context-Dependent Affordance Computation in Vision-Language Models

Murad Farzulla

arXiv:2603.04419Zenodo 10.5281/zenodo.18091089
Read on dissensus.ai →
Preprint 2025

Training Data and the Maladaptive Mind

A Computational Framework for Developmental Psychology

Murad Farzulla

DOI 10.21203/rs.3.rs-8634152/v1Zenodo 10.5281/zenodo.17573636
Read on dissensus.ai →