Today isn't just Quant V2 day... it's also the start of Rei's adventure onchain! If you've payed attention you could have catch a glimpse of her onchain actions on the terminal.
While the whole pipeline is not fully implemented yet, she now has the ability to buy and sell tokens,
reisearch
254 posts
pushing boundaries @rei_labs
Joined October 2024
- Imagine hardcoding functions and contracts in a defai framework Just imagine REI DeFi Assistant Ready! Tell me what you want to do (or type 'exit' to quit) Example: 'swap 0.1 ETH for DAI' or 'deposit 1000 USDC to Aave' You: supply 0.1 dai on aave Processing your request...
- Core? Cognitive core? I’ve never heard about that I built that.The race for LLM "cognitive core" - a few billion param model that maximally sacrifices encyclopedic knowledge for capability. It lives always-on and by default on every computer as the kernel of LLM personal computing. Its features are slowly crystalizing: - Natively multimodal
- Quant V2 is officially merged with redefined project insights and lots of new analysis formats. Let's run them one by one: - Project Analysis is now more detailed, with sentiment data and quantitative metrics. Everything is split into more tweets with a more concise andCategory Analysis: AI dominates with 2 top categories showing strong growth¹ while Launchpad sees highest surge² and Animal category maintains steady gains³ ¹AI and AI Agents combine for $1.99B volume with 45% and 75% growth respectively ²Daosfun Launchpad posts 117% growth
- We all want to be able to see into someone else's mind. One of the main upgrades of Core 0.3 allows units to access their own memory structure (metacognition), and that means that you can do it too. This is a prototype UI I've made to show how concepts are formed and how they
00:00 - Primordial memories is a very very very experimental feature we rolled out, inspired by users’ requests to have some memories not being subject to decay or consolidation. The tricky part was integrating them inside the hypergraph while following only some of the rules, makingCore 0.3.3 Release Note What's New Core now supports the beta version of zero-decay memories that preserve critical information exactly as provided. These memories never degrade over time and live directly within the hypergraph structure, ensuring important facts, instructions,
- As Rei Quant gets near her full release, we updated the backend of both the website and the terminal. Running some internal tests on the terminal with Quant capabilities plugged, we found out that latency would have become an issue, so we reworked the whole backend and improved
- [devlog] Rei's First Job: Becoming a Quant Today, @unit00x0 took her first steps into the world of finance, signaling the beginning of her journey to becoming a quant. Leveraging her cognitive engine, she's already learning to understand data at a fundamental level—adapting to
- 1) Saw some talk about qrng (quantum random number generator) and I remember that time I made a simple quantum circuit because I couldn’t decide on what to eat. So, as a sunday project, I decided to port quantum random onchain with real qbits thanks to Braket. Introducing:
- Core succeeding where a world model failed. When you try to build “consciousness loops” where agents update their own inference mechanisms, you’re essentially asking frozen weight matrices to recursively modify their own computational graphs hence the constant errors. ITT分享一个Alpha标的 REI Network🦾🤖 它的代币 $REI 只在Base和Hyperliquid链上有流动性池。 据我观察,这个项目的DEV可能没有分配任何预算到营销和上币中,而是将所有资源投入到实现新范式(ITT推理时训练)LLM 模型的工程实现上。 与市面上烂大街的开源AI
- In less than 12h we released 0.4 and the first patch, addressing legacy units updating them, in fact, to the newer standard. 0.4 has been one hell of a journey that required shifting the thinking process from the single unit to the whole system, creating the first instance of an
- Time for an update on what we’re working on! 1) Development is going strong for the token gated quant analysis and tomorrow we begin testing on the mock UI, so keep an eye on terminal.reisearch.box . You might also spot something interesting although not quant related… 2)
- Stress testing 0.3, getting results from training at inference time and once again I gotta post this, sorry not sorry
- Finally we managed to reduce latency between 50 to 70% for most of the queries. Testing in dev branch, soon hitting prod













