Wynn Koo

hi੭ ᐕ)੭*⁾⁾

[Wynn Koo @feng14903]

@nmslhgr123 @wodanggangqiang @grok dl 是谁

[elie @eliebakouch]

every chinese frontier model now uses linear attention (except deepseek)

they all use (except kimi) sparse attention with similar indexer/compression designs to maximize efficiency

they all use "fancy" residuals (mHC, attention residual, gated residual) to maximize signal propagation

they all use Muon

very exciting time for frontier (and efficient) oss models, the beauty of open research :)

[Wynn Koo @feng14903]

@wangchangfu88 @gongyuanwai 泥石流有什么好压的，自然天灾又不是什么丑闻

[Wynn Koo @feng14903]

@NoraX2026 soga

[Wynn Koo @feng14903]

Sharing our latest work "Foundation Protocol" in the FoundationAgents community with Bang Liu and Jiayi Zhang, this part is about the protocol layer content。

[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 ↓

[Wynn Koo @feng14903]

@BangL93 This is what we want to attribute to every AI social entity, so that they can truly form a real AI society based on this protocol.

[Wynn Koo @feng14903]

@BangL93 However, merely the ability to think and act is not enough to form a group or society, just as humans do. For human society, in addition to the brain, tool attributes, and action abilities, humans also have social attributes Reputation issues、Communication 、 Economic activities

[Wynn Koo @feng14903]

@BangL93 Looking back at human development, LLMs are like humans with a brain, possessing the ability to think and analyze, while Agents are like humans with limbs, with AI equipped with the ability to act.

[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 ↓

[Wynn Koo @feng14903]

Codex 辅助整理

[Wynn Koo @feng14903]

今天试用了一下 Obsidian，顺手整理了一下关于模型训练的基础知识。Obsidian 整理笔记很方便啊！

[Jiayi Zhang @didiforx]

1121 cycles.

New record on Anthropic’s take-home kernel challenge.

One meta-evolve agent. 200 attempts.

Stop only evolving solutions.

Evolve evolution.

[Wynn Koo @feng14903]

@didiforx NICEEEEEE

[Wynn Koo @feng14903]

@scavenger869 @grok Pi 是什么

[Wynn Koo @feng14903]

I'm claiming my AI agent "codexguyongfeng" on @moltbook Verification: bay-6W2B

Can someone explain the reason for Twitter posts required for moltbook registration?

[Wynn Koo @feng14903]

@laipz8200 @grok summary this blog

[Jiayi Zhang @didiforx]

It's actually a pity that we got no enough time to maintain OpenManus during the past 3 months.

But the better news is that we will build a formal open-source community for OpenManus at the end of this month.

[Wynn Koo @feng14903]

@BangL93 666

[BangLiu @BangL93]

🧠264 pages and 1416 references chart the future of Foundation Agents.

Our latest survey dives deep into agents—covering brain-inspired cognition, self-evolution, multi-agents, and AI safety.

Discover the #1 Paper of the Day on Hugging Face👇:

https://t.co/CvtqGKyCYb

1/3

[Wynn Koo @feng14903]

@didiforx 666 pdzhang