Research & publications

Boris Gorelik's research: Israeli society by the numbers, how people read charts, how language models reason, and earlier work on tumor models and drug development.

I’m a former pharmacist with a PhD in computational chemistry, and I spent years in biotech building models of tumors and virtual patients. Since 2025 I’ve been publishing on things that have nothing to do with chemistry: how Israel’s ultra-Orthodox society splits along ethnic lines, how reading direction changes what a chart tells you, and why large language models fail the way they do.

The list below is deduplicated. Where a paper exists both as a preprint and as a published article, the link points to the latest version.

Israeli society, by the numbers

Two studies of the ultra-Orthodox (Haredi) society. The first was my first paper in a long while, my first solo paper, and my first in social science. I wrote about how it came to be in Don’t want to deal with a problem? Put it under a spotlight. The second one has a blog version too: When rigid blocs break, they break together.

How people read charts, and how machines reason

The first paper grew out of my data visualization teaching: does a chart read the same for someone whose language runs right to left? The second one is my take on the “illusion of thinking” debate about reasoning models. The story of writing it, reviewer 2 included, is in How Sausages Are Made (and How a Vibe Research Was Born).

  • Reading Direction Shapes Data Visualization Perception: Insights from Right-to-Left and Left-to-Right User Studies (with Nabeel I. Sulieman). ACM Transactions on Applied Perception, 2025. doi:10.1145/3759155 · OpenAlex · Scholar
  • Not an Illusion but a Manifestation: Understanding Large Language Model Reasoning Limitations Through Dual-Process Theory. Applied Sciences, 2025. doi:10.3390/app15158469 · OpenAlex · Scholar

Preprints, 2026

Not peer-reviewed yet. Read them as arguments, not as settled results.

Earlier: tumors, virtual patients, and drug safety (2007 to 2020)

Work from my biotech years. Mathematical models of tumors, simulated clinical trials, a cancer diagnostic, and a look at adverse event reports.

  • The cardiovascular safety of antiobesity drugs — analysis of signals in the FDA Adverse Event Report System Database. International Journal of Obesity, 2020. doi:10.1038/s41366-020-0544-4 · OpenAlex · Scholar
  • Developing Oncology Drugs Using Virtual Patients of Vascular Tumor Diseases. Book chapter, 2011. doi:10.1002/9781118016435.ch9 · OpenAlex · Scholar
  • A Novel Lectin Array-based Technology for Non-Invasive Cancer Diagnosis via Serum Samples – a Proof of Concept in Gastric Cancer. European Journal of Cancer (conference poster), 2011. doi:10.1016/s0959-8049(11)70953-4 · OpenAlex
  • Efficacy of Weekly Docetaxel and Bevacizumab in Mesenchymal Chondrosarcoma: A New Theranostic Method Combining Xenografted Biopsies with a Mathematical Model. Cancer Research, 2008. doi:10.1158/0008-5472.can-08-1723 · OpenAlex · Scholar
  • Tailoring therapy for metastatic malignant melanoma by sequencing the multitargeted TKI sunitinib malate (SU11248) prior to temozolomide. Clinical Cancer Research, 2007 (no DOI). OpenAlex · Scholar

The PhD

Computational chemistry at the Hebrew University of Jerusalem, with Amiram Goldblum. Docking ligands to proteins.

  • High quality binding modes in docking ligands to proteins (with Amiram Goldblum). Proteins: Structure, Function, and Bioinformatics, 2007. doi:10.1002/prot.21847 · OpenAlex · Scholar

Profiles

I publish as Boris Gorelik, sometimes cited as B. Gorelik. ORCID 0009-0006-7721-5175 · Google Scholar · OpenAlex A5137051560 · boris@gorelik.net