I'm Omar — an ML engineer and interpretability researcher in New York City. I build instruments for seeing inside models.
Currently building DinoDreamer — a feature-visualization instrument for navigating DINOv2's representation space. Before that: SubspaceLens , measuring subspace geometry in diffusion transformers with MCR² instead of sparse autoencoders, and Latent Dance , a music visualizer where audio re-steers SDXL through 20,480 SAE features every frame — a live control loop at 50 FPS, not batch generation.
NYU CS (M.S. '24), Recurse Center alum, two peer-reviewed computer-vision publications. Open to ML engineering and research engineering roles — say hi .
Work Time:
____ ____ _ ____ ____ _ ____ ____ / __ \ / __ \ (_) / __ \ / __ \ (_) / __ \ / __ \ / / / / / / / / _ / / / / / / / / _ / / / / / / / / / /_/ / / /_/ / (_) / /_/ / / /_/ / (_) / /_/ / / /_/ / \____/ \____/ \____/ \____/ \____/ \____/
Break Time:
____ ____ _ ____ ____ _ ____ ____ / __ \ / __ \ (_) / __ \ / __ \ (_) / __ \ / __ \ / / / / / / / / _ / / / / / / / / _ / / / / / / / / / /_/ / / /_/ / (_) / /_/ / / /_/ / (_) / /_/ / / /_/ / \____/ \____/ \____/ \____/ \____/ \____/
_______________ |*\_/*|________
| ___________ | .-. .-. ||_/-\_|______ |
| | | | .****. .****. | | | |
| | 0 0 | | .*****.*****. | | 0 0 | |
| | - | | .*********. | | - | |
| | \___/ | | .*******. | | \___/ | |
| |___ ___| | .*****. | |___________| |
|_____|\_/|_____| .***. |_______________|
_|__|/ \|_|_............*..............._|________|_
/ ********** \ / ********** \
/ ************ \ / ************ \
-------------------- --------------------State: WORKING