- Szolgáltatás
Iván Arcuschin Moreno
Independent researcher specializing in AI Safety and Mechanistic Interpretability.
Bemutatkozás
Iván Arcuschin Moreno is an Independent Researcher working on AI Safety. He earned a Computer Science PhD from the University of Buenos Aires, Argentina, specializing in Automated Test Generation for Android apps, with research published in top Software Engineering conferences like ICST and AST. He participated in the ML Alignment & Theory Scholars (MATS) Program, working on AI Safety, including developing InterpBench, a benchmark for evaluating Mechanistic Interpretability techniques, published at NeurIPS 2024. Current work includes research on Chain-of-Thought Reasoning in LLMs regarding subtle unfaithfulness, where models construct superficially coherent yet contradictory arguments. Research areas span Mechanistic Interpretability, LLM reasoning faithfulness, and Automated Test Generation.
Szolgáltatások, tevékenységek
- Independent research on AI Safety
- Automated Test Generation for Android apps research
- Participation in MATS Program
- Development of InterpBench benchmark
- Research on Chain-of-Thought Reasoning unfaithfulness in LLMs