Zoe De Simone

Zoe De Simone

PhD in Computer Science at CSAIL · NSF Graduate Fellow

Zoe De Simone is a PhD student in Computer Science at MIT and an NSF Graduate Research Fellow. Her research focuses on human-AI collaboration and alignment in generative AI systems.

She previously completed dual master's degrees in Building Technology and EECS at MIT, where she built decision-support tools for equitable city-scale building decarbonization, and earned her B.Arch from Cornell, where she co-developed Eddy3D and worked on surrogate ML models for airflow simulation.

She has collaborated with Ashia Wilson, Mitchell Gordon, Fredo Durand, Arvind Satyanarayan, Christoph Reinhart, Timur Dogan, and Saleh Kalantari, and worked in industry at Point72 and SmartFab on applied AI systems in finance and manufacturing.

Email · Google Scholar · GitHub · LinkedIn

Previously: Sustainable Design Lab (MIT), Cornell Environmental Systems Lab, Foster + Partners R&D, SmartFab, Point72.

zoed@mit.edu

Publications & Projects

DiffusionWorldViewer

DiffusionWorldViewer

Zoe De Simone, Angie Boggust, Arvind Satyanarayan, Ashia Wilson

arXiv (2023)

Interactive web system for analyzing and editing bias in text-to-image diffusion models.

Paper Code Demo
City-Scale Decarbonization Toolkit

City-Scale Decarbonization Toolkit

Zoe De Simone, Christoph Reinhart, Mariana Arcaya

Energy Policy (2025) · MIT MS Thesis (2024)

Human-in-the-loop stochastic optimization framework for equitable retrofit planning.

Paper Thesis
Eddy3D Indoor Airflow Simulation

Eddy3D — Indoor Airflow Simulation Tool for CAD

Zoe De Simone, Patrick Kastner, Timur Dogan

Building Simulation Conference (2021)

Development of the Indoor Eddy3D simulation tool for floor plan layouts and ventilation impact on indoor air quality and pathogen spread. 20k+ downloads.

Paper Software
CFD Surrogate: Surrogate vs baseline CFD

Real-Time Interactive Airflow Feedback Using CFD Surrogates

Zoe De Simone, Patrick Kastner, Timur Dogan

Forthcoming · Cornell Environmental Systems Lab

Physics-informed ML surrogate model reducing CFD simulation time from 3+ hours to ~1 second. Enables real-time, interactive airflow feedback from input encodings for early-stage design exploration.

Paper forthcoming Context
Creo progressive co-creative ideation pipeline

Creo: From One-Shot Image Generation to Progressive, Co-Creative Ideation

Zoe De Simone, Angie Boggust, Fredo Durand, Ashia Wilson, Arvind Satyanarayan

arXiv (2026)

Creo is a multi-stage text-to-image system that supports ideation from sketch-like abstractions to high-fidelity outputs with stage-level edits and locking. In comparative studies, people reported stronger ownership and produced more diverse results than with one-shot generation.

Paper
Alignment has a Fantasia Problem interaction diagram

Alignment has a Fantasia Problem

Nathanael Jo, Zoe De Simone, Mitchell Gordon, Ashia Wilson

arXiv (2026) · ICLR HCAIR Workshop

This work argues that instruction-tuned systems often assume prompts are complete expressions of intent even when users are still forming goals. It characterizes these breakdowns as Fantasia interactions and proposes alignment approaches that support intent formation over time through cognitively informed interaction design.

Paper ICLR HCAIR Workshop
Window-to-Wall Ratio Detection using SegFormer

Window-to-Wall Ratio Detection using SegFormer

Zoe De Simone, Sumit Biswas, Oliver Wu

CVPR Workshop on Computer Vision in the Built Environment (2024)

Semantic segmentation pipeline for façade glazing ratio prediction from street-view imagery.

Paper Code
FlexiArch

FlexiArch

Zoe De Simone, Tx Zheng, Bill Xu, Saleh Kalantari

Journal of Building Engineering (2023)

Computational tool to assess building longevity through flexibility; practitioner survey (n=237).

Paper

Full publication list on Google Scholar →

Awards and Honors

Selected talks

News and Features