𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-'

Creator of Mnemon @mnemondev | DeFi analyst | Agentic AI | Alumni @ZJU_China | Prev. @BytedanceTalk | $HYPE $PENDLE

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

GitHub ID: Grivn

项目 mnemon：https://t.co/VQ2qYRSvjt
LLM-supervised persistent memory for AI agents.
支持一键接入任何实现了标准 hook 的 ai agent。

目前正在基于 mnemon 的设计思想，实现记忆驱动的 mnemon-harness 框架，目前处于实验阶段。

其他相关介绍：
https://t.co/cZS0fbvzKt
https://t.co/9bTbJA94dA

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

As the creator of Mnemon, it’s an honor to see Dr. Vivian Balakrishnan (@VivianBala) exploring it as part of his personal AI stack for persistent knowledge and second-brain workflows.

Mnemon started as a memory system, but our long-term direction is a memory-driven harness layer for durable and self-evolving agents.

We believe future agents need more than context windows. They need persistent state, evolution loops, orchestration, and runtime continuity around the host agent.

Our goal is simple: make it possible for any agent to plug into Mnemon and gain persistent cognition.

Still early. Still building.

Feel free to try it out and explore the future of persistent agents with us.

https://t.co/VQ2qYRRXtV

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@ryandcrypto Hi. Connect with $HYPE .

Hyperliquid.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@coinbureau cliché

[Hunter Biden @HunterBiden]

@gmoneyNFT Hyperliquid

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@HunterBiden @dongdongRobin @gmoneyNFT Hyperliquid.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@holegots @OpenAI @gdb https://t.co/hoDsxM2YCD

[Greg Brockman @gdb]

rust is a perfect language for agents, given that if it compiles it's ~correct

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@holegots @OpenAI @gdb Hey, here’s a guy with some opinions.

[Omar @TheOneandOmsy]

Jeff Sprecher, founder and CEO of $ICE (owns the NYSE) on Hyperliquid:
"This Hyperliquid that we're talking -- if you haven't heard about it, it's bigger than NASDAQ, okay? It's 11 people. You look at it, you're like, wow, that's pretty something."

If it wasn't clear before, hyperliquid:native has grown far beyond crypto. The incumbents have noticed, are paying close attention, and even spending time with the team

Bernstein excerpt below and worth the read imo:

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@Hypurr Hyperliquid

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@Gavriel_Cohen @vercel @Docker @mondaydotcom @NanoClaw_AI Congratulations!

[Dylan Zhang @dylan_works_]

Wrote up something fun I’ve been poking at: when LLM agents repeatedly rewrite their own experiences into textual “lessons,” their memory can get worse, not better.

Across several environments, we found a recurring pattern: forced consolidation often degrades useful experience into faulty or overgeneralized memories.
Interestingly, models seem much better at managing examples as memory objects than at distilling them into reusable routines.

Maybe we should be more careful about asking agents to constantly “consolidate” experience into lessons 🤔?

I’m new to this area, so I’d love thoughts. I may be missing context or just wrong on parts of it — please don’t hesitate to let me know! Discussions are always welcome.

https://t.co/0IiQvOJxQQ

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@SherryYanJiang @VivianBala hopefully soon 👀 would love to meet in singapore sometime.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

As the creator of Mnemon, it’s an honor to see Dr. Vivian Balakrishnan (@VivianBala) exploring it as part of his personal AI stack for persistent knowledge and second-brain workflows.

Mnemon started as a memory system, but our long-term direction is a memory-driven harness layer for durable and self-evolving agents.

We believe future agents need more than context windows. They need persistent state, evolution loops, orchestration, and runtime continuity around the host agent.

Our goal is simple: make it possible for any agent to plug into Mnemon and gain persistent cognition.

Still early. Still building.

Feel free to try it out and explore the future of persistent agents with us.

https://t.co/VQ2qYRRXtV

[AI Engineer @aiDotEngineer]

"You cannot govern a technology you have only been briefed on."

Singapore Minister for Foreign Affairs, Dr. @VivianBala, echoing @karpathy and @yacineMTB on why he runs NanoClaw: "you can outsource memory and computation, but you cannot outsource your understanding"

https://t.co/z4Aidf89ha

He also shared his tech stack for running his second brain for Singapore's Foreign Affairs Ministry and parliamentary affairs:
- @AnthropicAI Claude Agent SDK
- Baileys + WhatsApp
- Mnemon (Graph Memory)
- @ollama + @nomic_ai
- @ggerganov Whisper.cpp + OneCLI

With special notes on how he handles security and isolation, and what implications he sees for Singapore Inc.

[Louis Maddox @permutans]

mnemon https://t.co/2JKLVgfPsk

> Most memory tools embed own LLM inside the pipeline. [In mnemon] your host LLM is the supervisor… handles deterministic computation (storage, graph indexing, search, decay)… makes judgment calls (what to remember, how to link, when to forget).

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@dongdongRobin @VivianBala 🫡

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@0xHuu @VivianBala Job's not finished.

Hyperliquid.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@slap__tjips @VivianBala Already being explored.

Hermes currently doesn’t expose explicit runtime hooks like Codex, Claude Code, OpenClaw, or NanoClaw, so proper integration needs a slightly different harness approach. We’ll introduce it once the design is ready.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@zeroxkyle @VivianBala Mnemon is built for easy integration with any host agent. And it looks like it’s really starting to work.

[rachael de foe @unprofeshme]

“you cannot govern a technology you’ve only been briefed on.”

this morning minister vivian balakrishnan walked into a room full of builders and did something that genuinely mattered:
he showed up as one himself.

he opened up his actual workflow, talked through the second brain he’s been building, the tradeoffs, the implications, the limits, and the things he’s still figuring out.

and I’m sure, before he spoke, there were skeptics in the room.

builders have heard enough people talk about technology from a distance to know the difference between someone who has touched the tools and someone who has only read summaries about them.

what struck me most wasn’t even the talk itself.

it was what happened after.

he stayed.

he went off stage and spent time with young builders. students. engineers. people still figuring out their place in all of this.
then he stayed in the room to watch @Gavriel_Cohen’s @NanoClaw_AI talk because he genuinely wanted to hear it.

and as someone on the organising team, I can tell you that none of it was staged and all of it was him.

that detail will stay with me for a long time.

there were students sitting in the front row today who now have a completely different reference point for what leadership can look like.

he wasn’t speaking above the work, he was actually doing it.

I’m so aware that what happened today was much bigger than us.

thank you, minister @vivianbala

putting together @65labslah @aiDotEngineer has always been rooted in a hope that singapore can become a place where serious builders collide openly, generously and ambitiously.

today, you helped us show that it is.

[𝗚𝗥𝗜𝗩𝗡.𝗛𝗟 🌎-' @grivn_eth]

@aiDotEngineer @VivianBala @karpathy @yacineMTB Thanks a lot 🙏