BangLiu

I am an associate professor at University of Montreal and Mila.

[BangLiu @BangL93]

check our blog here: https://t.co/W0kGn5gkXn

[BangLiu @BangL93]

Agents are moving from tool use to paid work.
But paid work needs more than wallets.
It needs contracts, evidence, settlement, and reputation.

That’s Foundation Protocol:
the accountability layer for the agent economy.

Watch the feature overview ↓

[BangLiu @BangL93]

🤗：https://t.co/JpppVnkjcZ
📄：https://t.co/ykKc3Ih3tx
💻Founda Protocol：https://t.co/lzfeKnJUDf
💻Demo：https://t.co/ztQ6uFQZmW

[BangLiu @BangL93]

@MingchenZhuge @tydsh @karpathy @DrJimFan @SchmidhuberAI @_akhaliq @hardmaru @zechunliu @YoungXiong1 @HaoZhe65347 @cai_zhipeng Amazing work！

[Mingchen Zhuge @MingchenZhuge]

🫱 Introducing 𝐍𝐞𝐮𝐫𝐚𝐥 𝐂𝐨𝐦𝐩𝐮𝐭𝐞𝐫s:

𝐰𝐡𝐚𝐭 𝐢𝐟 𝐀𝐈 𝐝𝐨𝐞𝐬 𝐧𝐨𝐭 𝐣𝐮𝐬𝐭 𝐮𝐬𝐞 𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐫𝐬 𝐛𝐞𝐭𝐭𝐞𝐫, 𝐛𝐮𝐭 𝐛𝐞𝐠𝐢𝐧𝐬 𝐭𝐨 𝐛𝐞𝐜𝐨𝐦𝐞 𝐭𝐡𝐞 𝐫𝐮𝐧𝐧𝐢𝐧𝐠 𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐫 𝐢𝐭𝐬𝐞𝐥𝐟?

Beyond today's conventional computers, agents, and world models, Neural Computers (NCs) are new frontiers where computation, memory, and I/O move into a learned runtime state.

We ask: whether parts of runtime can move inward into the learning system itself. This is our first step toward the Completely Neural Computer (CNC): a general-purpose neural computer with stable execution, explicit reprogramming, and durable capability reuse.

Work done with Mingchen Zhuge (@MingchenZhuge), Changsheng Zhao, Haozhe Liu (@HaoZhe65347 ), Zijian Zhou (@ZijianZhou524 ), Shuming Liu (@shuming96 ), Wenyi Wang (@Wenyi_AI_Wang ), Ernie Chang (@erniecyc ), Gael Le Lan, Junjie Fei, Wenxuan Zhang, Zhipeng Cai (@cai_zhipeng ), Zechun Liu (@zechunliu ), Yunyang Xiong (@YoungXiong1 ), Yining Yang, Yuandong Tian (@tydsh ), Yangyang Shi, Vikas Chandra (@vikasc), Juergen Schmidhuber (@SchmidhuberAI)

[BangLiu @BangL93]

Given a fixed budget: scale out with more agents, or scale up a single one?
We derive a sharp criterion comparing the "organization exponent" to the single-agent scaling exponent to determine when multi-agent expansion wins and when it doesn't.

📑 arXiv: https://t.co/q1pQN4MSYj

[BangLiu @BangL93]

structural prototype construction and multi-objective–driven optimization.

The framework remains stable and controllable under multiple continuous property constraints, and achieves clear improvements over strong baselines on both drug discovery and materials design tasks.

[BangLiu @BangL93]

📄
In collaboration with Samsung, we propose a controllable, multi-stage molecular generation framework that turns molecular generation from a one-shot black box into two explicit stages.
👉 arXiv: https://t.co/UoH6hRRDTB
🤗 https://t.co/TCpOgP3yX0

[BangLiu @BangL93]

Takeaway: if we want agents that truly improve over time, we need skill lifecycle management—debugging, freezing, and refactoring as first-class learning operators.

[BangLiu @BangL93]

Glad to share our new work “Evolving Programmatic Skill Networks (PSN)”
Arxiv: https://t.co/1Jca840qcJ
HuggingFace: https://t.co/6XDEjA7UDS

We treat skills as executable code with explicit pre/post-conditions, and the skill library is wired into a compositional evolving graph

[BangLiu @BangL93]

- REFLECT: trace-based fault localization / credit assignment over the executed subgraph (backprop-like)
- Maturity-aware update gating: progressively stabilizes reliable skills to reduce forgetting
- Online refactoring + rollback validation:keep the network compact

[BangLiu @BangL93]

I’m excited to share that I’ve received an Amazon Research Award for my proposal Foundation Agents and Protocol for Collaborative Agentic AI at University of Montreal. Learn more about the program on the @AmazonScience website: https://t.co/tFxoePQYbr
#AmazonResearchAwards

[BangLiu @BangL93]

Glad to share our recent research on ReCode: Unify Plan and Action

[Zhaoyang Yu @ZhaoyangYu22356]

Is ReAct the final for LLM Agents? 🤔
Not anymore. ReAct is stuck step-by-step. Planning agents separate plans from actions.
Introducing ReCode: agents control their decision granularity, like humans do.
20.9% better, 78.9% cheaper, 3.7× data efficient. The ReAct era is over.
1/5

[Mila - Institut québécois d'IA @Mila_Quebec]

Mila's annual supervision request process is now open to receive MSc and PhD applications for Fall 2026 admission! For more information, visit https://t.co/r01eLcY1P4

[BangLiu @BangL93]

@xie_yupeng87927 Thanks for liking it😊

[BangLiu @BangL93]

🤖Check The Hitchhiker’s Guide to Agents HERE🤖

Our Foundation Agents Survey V2 level up to 396 pages – every chapter is a full-on survey itself!
🧠 Agent Framework & Components
🌍 World Model & Memory
🔄 Self-Evolution
👥 Multi Agents
🛡️ Safety

1/4

[BangLiu @BangL93]

@xinyu_liu91723 Thanks for liking it😊

[BangLiu @BangL93]

@F_A_Org Join us @F_A_Org to build the future of AI agents together!

[BangLiu @BangL93]

@oliverlee1999 Thank you!

[BangLiu @BangL93]

@oliverlee1999 😊 Thank you!

[BangLiu @BangL93]

@yuchen1971 Thank you!😊