iiinova — where your digital self keeps evolving

Mason Nakamura

@masonnaka.bsky.social

Thinking about value alignment, RL, multi-agents, and embodied agents www.masonnakamura.com

Mason Nakamura @masonnaka.bsky.social · 30/10/2025

This work was done in collaboration with @akumar2709, @saad-ai.bsky.social, Sahar Abdelnabi, Shlomo Zilberstein, and @ebagdasa.bsky.social. 📄paper: arxiv.org/pdf/2510.14312 💻Code: github.com/umass-aisec/... 🌐Project Website: aisec.cs.umass.edu/projects/ter...

arxiv.org

000

Mason Nakamura @masonnaka.bsky.social · 30/10/2025

Our attack evaluation covers confidentiality (info leakage), integrity (adversarial agent & comm‑poisoning), availability (context overflow). Additionally, we integrate 3 cooperative DCOP environments: 📅 MeetingScheduling, 🏡 SmartGrid, and 🧎PersonalAssistant.

100

Mason Nakamura @masonnaka.bsky.social · 30/10/2025

Why? MASs amplify capability and risks; private data + cross‑agent comms create large attack surfaces. Terrarium provides a controllable, observable sandbox that uses MCP servers and agent-to-agent comm via blackboards for reproducible studies on a new agent paradigm.

100

Mason Nakamura @masonnaka.bsky.social · 30/10/2025

🚨New preprint: Terrarium-an open source, blackboard-based testbed for studying safety, privacy & security in LLM multi‑agent systems (MAS). We showcase the vulnerabilities and safety considerations of agentic MASs in this modular and configurable framework. 🧵 #AISafety #LLMAgents #Agents

150