# Research

Canonical: https://www.spencerhong.com/work/research

## Research

I study complex systems through noisy data, inaccurate models, and unrealistic simulations. My work ranges from density functional theory & molecular dynamics for chemical and molecular systems to networks¹ & multimodal models for biological and social systems.

Just like the Human Genome Project, I’m motivated by finding and scaling noisy data to make our models ever so slightly more accurate, and our simulations a little closer to the world we’re trying to explain.² And if we can explain it, can we use that understanding to build something new?

My work has been covered by The New York Times, Wall Street Journal, OpenAI, and others.

1. not scale-free

2. I’m motivated by The Bitter Lesson, but I read R. A. Fisher’s Gene Frequencies in a Cline Determined by Selection and Diffusion first. I also acknowledge Fisher’s involvement in scientific racism and eugenics, views I completely disavow.

- [not scale-free](https://arxiv.org/pdf/physics/0510216)
- [The Bitter Lesson](https://www.cs.utexas.edu/~eunsol/courses/data/bitter_lesson.pdf)
- [Gene Frequencies in a Cline Determined by Selection and Diffusion](https://pubmed.ncbi.nlm.nih.gov/14791572/)

## Selected papers

- [A digital archive reveals how a funding agency cooperated with academics to support the nascent field of genomics — Nature Communications, 2026](https://www.nature.com/articles/s41467-026-71700-9)
- [EPSVec: Efficient and Private Synthetic Data Generation via Dataset Vectors — ICML, 2026](https://proceedings.mlr.press/v306/banayeeanzade26a.html)
- [CAGE: Coherence-Aware Graph Encoding for Retrieval-Augmented Generation — EMNLP, 2026](https://arxiv.org/abs/2609.04647)
- [RAFFLES: Reasoning-based Attribution of Faults for LLM Systems — EACL, 2026](https://aclanthology.org/2026.eacl-long.359/)
- [DF-RAG: Query-Aware Diversity for Retrieval-Augmented Generation — Findings of EACL, 2026](https://aclanthology.org/2026.findings-eacl.150/)
- [SAFARI: Scaling Long Horizon Agentic Fault Attribution via Active Investigation — ICML AIWILD Workshop, 2026](https://arxiv.org/abs/2606.24626)
- [BuyTheBy: A dataset of 18,710 text-based paper mill advertisements with 51,812 timestamped prices — arXiv preprint, 2026](https://arxiv.org/abs/2604.24576)
- [The entities enabling scientific fraud at scale are large, resilient, and growing rapidly — PNAS, 2025](https://doi.org/10.1073/pnas.2420092122)
- [Widespread misidentification of scanning electron microscope instruments in the peer-reviewed materials science and engineering literature — PLOS ONE, 2025](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0326754)
- [Exploitation of intellectual property systems for the manipulation of academic reputations — International Journal for Educational Integrity, 2025](https://link.springer.com/article/10.1007/s40979-025-00185-8)
- [Ethical considerations in utilizing artificial intelligence for analyzing the NHGRI’s History of Genomics and Human Genome Project archives — Journal of eScience Librarianship, 2024](https://doi.org/10.7191/jeslib.811)
- [How well do implicit solvation models represent intermolecular binding energies in organic-inorganic solutions? — Computational Materials Science, 2019](https://doi.org/10.1016/j.commatsci.2019.109138)

## Navigate

- [Overview](https://www.spencerhong.com/)
- [Experience](https://www.spencerhong.com/work/experience)
- [Music](https://www.spencerhong.com/music)
- [About](https://www.spencerhong.com/about)
- [Contact](https://www.spencerhong.com/contact)
- [Privacy](https://www.spencerhong.com/privacy)
- [Blog](https://www.spencerhong.com/blog)
- [On Agentic Memory](https://www.spencerhong.com/blog/on-agentic-memory)
- [AI agents](https://www.spencerhong.com/agents)