Lattice AI Research

Research focus

Red-Teaming & Adversarial Robustness (2)Constitutional AI & AI Ethics (1)Data Curation & Synthetic Data (1)Code Generation & Program Synthesis (1)

Frequent co-authors

Shunan Zhu (1)Yonghao Yu (1)H. Ochiai (1)Haochen Liang (1)

Papers (2)

Apr 8, 2026

3w ago

FedDetox: Robust Federated SLM Alignment via On-Device Data Sanitization

Turn toxic federated data into a safety win: FedDetox uses on-device sanitization to transform potential data poisoning into positive safety signals for small language models.

Shunan Zhu, Yonghao Yu, H. Ochiai +1

Constitutional AI & AI Ethics Data Curation & Synthetic Data Red-Teaming & Adversarial Robustness

Mar 1, 2026

Great Bay UniversityMar 1, 2026·also UTokyo

NeuroSCA: Neuro-Symbolic Constraint Abstraction for Smart Contract Hybrid Fuzzing

LLMs can surgically remove semantic noise from smart contract constraints, dramatically accelerating hybrid fuzzing without sacrificing soundness.

Haochen Liang, Hideya Ochiai

Code Generation & Program Synthesis Red-Teaming & Adversarial Robustness

Search

Hideya Ochiai

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)