Scholar
Iván Arcuschin
Google Scholar ID: vfB-YhcAAAAJ
Independent Researcher
Citations & Impact
Citations
101
H-index
6
i10-index
1
Publications
14
Co-authors
5
list available
Publications
Resume (English only)
Academic Achievements
- Published papers in top Software Engineering conferences such as ICST and AST; published several papers including 'MIB: A Mechanistic Interpretability Benchmark', 'Chain-of-Thought Reasoning In The Wild Is Not Always Faithful', 'Understanding Reasoning in Thinking Language Models via Steering Vectors', 'InterpBench: Semi-Synthetic Transformers for Evaluating Mechanistic Interpretability Techniques'.
Research Experience
- Participated in the ML Alignment & Theory Scholars (MATS) Program, working on AI Safety; worked on InterpBench in 2024, a benchmark of semi-synthetic transformers with known circuits for evaluating Mechanistic Interpretability techniques; in 2025, worked on the paper 'Chain-of-Thought Reasoning in the Wild is Not Always Faithful'.
Education
- PhD in Computer Science, University of Buenos Aires, Argentina, Specialization: Automated Test Generation for Android apps (Advisor information not provided)
Background
- Independent Researcher, focusing on AI Safety. Earned a PhD in Computer Science from the University of Buenos Aires, Argentina, specializing in Automated Test Generation for Android apps.
