Leo de Castro

Researcher, J.P. Morgan Chase

Joined

May 2024

Names

Leo de Castro (Preferred)

Emails

****@mit.edu (Confirmed)

****@gmail.com (Confirmed)

****@jpmchase.com (Confirmed)

Personal Links

DBLP

Career & Education History

Researcher

J.P. Morgan Chase (jpmchase.com)

2024 – Present

PhD student

Massachusetts Institute of Technology (mit.edu)

2019 – 2024

Advisors, Relations & Conflicts

PhD Advisor

Vinod Vaikuntanathan

2019 – 2024

Expertise

No areas of expertise listed

Publications

HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions
Harish Karthikeyan, Leo de Castro, Antigoni Polychroniadou, Manuela Veloso, Tucker Balch
- Submitted to ICLR 2026
- Readers: Everyone
HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions
Harish Karthikeyan, Leo de Castro, Antigoni Polychroniadou, Manuela Veloso, Tucker Balch
- GenAI in Finance Poster
- Readers: Everyone
AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration
Harish Karthikeyan, Yue Guo, Udari Madhushani Sehwag, Leo de Castro, Antigoni Polychroniadou, Leo Ardon, Sumitra Ganesh
- SEA @ NeurIPS 2025 Poster
- Readers: Everyone
HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions
Harish Karthikeyan, Leo de Castro, Antigoni Polychroniadou
- RegML 2025 Poster
- Readers: Everyone
AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration
Harish Karthikeyan, Yue Guo, Udari Madhushani Sehwag, Leo de Castro, Antigoni Polychroniadou, Leo Ardon, Sumitra Ganesh
- RegML 2025 Poster
- Readers: Everyone
EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption
Leo de Castro, Daniel Escudero, Adya Agrawal, Antigoni Polychroniadou, Manuela Veloso
- ICML 2025 poster
- Readers: Everyone
Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption
Leo de Castro, Antigoni Polychroniadou, Daniel Escudero
- SafeGenAi Poster
- Readers: Everyone

Leo de Castro

Names

Emails

Personal Links

Career & Education History

Advisors, Relations & Conflicts

Expertise

Publications

HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions

HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions

AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration

HashMark: Watermarking Tabular/Synthetic Data For Machine Learning Via Cryptographic Hash Functions

AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration

EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption

Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption

Co-Authors