Summary
Joel Dapello is a scientist and engineer with a cross-functional background in AI and biology and over ten years of experience focused on accelerating bio and neuro research with AI/ML methods. Joel prioritizes impact, follow through, clear communication, and responsible development of AI/ML methods.
Experience
Graepel & Bianco Labs, Altos Labs | Machine Learning Scientist
October 2022 - present
- Currently leading a cross functional team to develop an agentic AI platform for therapeutic target assessment and prioritization
- Founded and scaled the unimodality and multimodality foundation model program
- Led a cross functional team of wet lab biologists, bioinformaticians, and ML engineers to develop and scale RNAseq, Imaging, and multimodality foundation models for therapeutic discovery
DiCarlo Lab, MIT | PhD Researcher
August 2018 - December 2022
- Adversarial vulnerability in artificial and biological neural systems [1,2,3,5]
- Geometric analysis of information processing in artificial and biological neural systems [4,6]
MIT IBM Watson AI Lab, IBM | Research Intern
June 2021 - August 2021
- Out of domain generalization using Invariant Risk Minimization and Model Agnostic Meta Learning
Cox Lab, Harvard University | PhD Researcher
August 2017 - August 2018
- Convolutional neural networks for decoding information from biological neural systems [7]
- Information theoretic analysis of generalization in artificial neural networks [8]
BioBright, LLC | Founding Engineer
July 2014 - April 2017
- Developed an integrated platform for improving efficiency and reproducibility in biology research
- Prototyped an NLP voice based note taking and action triggering systems for wet lab scientists and a convolutional neural network based method for automated experiment tracking
Boyden Lab, MIT | Research Affiliate
March 2015 - March 2017
- Novel optical device design for recording and stimulation of neural activity [10]
Robinson Lab, Rice University | Research Assistant
June 2013 - December 2013
- Controlling multiple presynaptic inputs with optogenetics and spatial light manipulation [11]
Education
Harvard University, Cambridge MA
PhD, Applied Math, August 2017 - September 2022
Hampshire College, Amherst MA
BA, Cellular and Molecular Biology, August 2011 - May 2014
Selected Publications, Talks, and Posters
[1] Dapello, J.*, Kar, K.*, Schrimpf, M., Geary, R., Ferguson, M., Cox, D. D., DiCarlo, J. (2023) Aligning model and macaque inferior temporal cortex representations improves model-to-human behavioral alignment and adversarial robustness, ICLR 2023 (Oral)
[2] Dapello, J., (2022) What can the Primate Brain Teach Us about Robust Object Recognition? Keynote Talk, New Frontiers in Adversarial Machine Learning Workshop, ICML, 2022
[3] Guo, C., Lee, M., Leclerc, G., Dapello, J., Rao, R., Madry, A., Dicarlo, J. (2022) Adversarially trained neural representations are already as robust as biological neural representations, ICML, 2022
[4] Dapello, J.*, Feather, J.*, Le, H.*, Marques, T., Cox, D., McDermott, J., DiCarlo, J., Chung, S. (2021) Neural Population Geometry Reveals the Role of Stochasticity in Robust Perception, NeurIPS, 2021
[5] Dapello, J.*, Marques, T.*, Schrimpf, M., Geiger, F., Cox, D., DiCarlo, J. (2020), Simulating a Primary Visual Cortex at the Front of CNNs Improves Robustness to Image Perturbations. NeurIPS, 2020 (Spotlight)
[6] Chung, S.*, Dapello, J.*, Cohen, U., DiCarlo, J. J., Sompolinsky, H. (2020), Separable Manifold Geometry in Macaque Ventral Stream and DCNNs. Poster, COSYNE 2020.
[7] Guitchounts, G., Lotter, W., Dapello, J., Cox, D., (2020), Stable 3D head direction signals in the primary visual cortex, biorxiv, 2020.09. 04.283762
[8] Saxe, A. M., Bansal, Y., Dapello, J., Advani, M., Kolchinsky, A., Tracey, B. D., & Cox, D. D. (2018) On the Information Bottleneck Theory of Deep Learning, ICLR, 2018.
[9] Fracchia, C., Dapello, J.. (2016). DEF CON 24: Reverse engineering biomedical equipment for fun and open science. DEFCON24, Biohacking Village
[10] Rodriques, S., Marblestone, A., Scholvin, J., Dapello, J., Sarkar, D., Mankin, M., Gao, R., Wood, L., Boyden, E. (2016) Multiplexed neural recording along a single optical fiber via optical reflectometry, Journal of Biomedical Optics, Vol. 21, Issue 5, 057003 (May 2016)
[11] Avants, B., Murphy, D.,, Dapello, J., Robinson, J., (2015) NeuroPG: open source software for optical pattern generation and data acquisition, Frontiers in Neuroengineering 2015/3/2