Lattice AI Research

Research focus

Tool Use & Agents (2)Architecture Design (Transformers, SSMs, MoE) (1)Scalable Oversight & Alignment Theory (1)Eval Frameworks & Benchmarks (1)

Frequent co-authors

Zhe Yu (1)Zhiyuan He (1)Yuxuan Huang (1)Lee Ka Yiu (1)

Papers (2)

Apr 24, 2026

Zhe Yu +7Apr 24, 2026

From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company

Forget rigid multi-agent pipelines: this framework lets you build self-organizing AI "companies" that dynamically recruit talent and adapt to tasks on the fly.

Zhe Yu, YuQi Fu, Zhiyuan He +5

Architecture Design (Transformers, SSMs, MoE)Scalable Oversight & Alignment Theory Tool Use & Agents

Apr 2, 2026

Haoyu Luo +10Apr 2, 2026

ATBench: A Diverse and Realistic Trajectory Benchmark for Long-Horizon Agent Safety

Current LLM safety evaluations miss the mark: ATBench reveals how risks in realistic, multi-step agent interactions emerge over time, challenging even the strongest models.

Haoyu Luo, Yuejin Xie, YuQi Fu +8

Eval Frameworks & Benchmarks Red-Teaming & Adversarial Robustness Tool Use & Agents

Search

YuQi Fu

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (2)