7/ 🧾 Read the full paper here tinyurl.com/3hrjcrrj #DeFi #FinTech #Crypto #AI @MorphoLabs @aave @compoundfinance @gauntlet_xyz @chaoslabs @BlockAnaliticaA

Finally, more interactive reviews in public

Details on Season 2: Sats Campaign 💥 Details in the following tweet:

During a gold rush sell shovels

Spark has successfully allocated 100 million DAI in fresh liquidity from Maker via our Metamorpho vault to sUSDe/DAI and USDe/DAI markets on Morpho Blue. This liquidity, sourced from Maker’s Direct Deposit Module, is designed to provide unmatched liquidity at scale for sUSDe and USDe users who wish to increase their token exposure through leverage. In addition to SparkLend, this deployment marks the second liquidity allocation from Maker to DeFi opportunities curated by Spark and facilitated by the Endgame plan realm. As a Maker SubDAO, Spark aims to develop DeFi infrastructure with exceptional liquidity from Maker. This is achieved by directing DAI to unique opportunities that offer substantial benefits for users and potential upside for MakerDAO and Spark. Now, users can leverage Maker's liquidity to increase their exposure to sUSDe and USDe, with unparalleled liquidity, unlocking a range of fresh opportunities in the DeFi market.

Leveraged USDe on Morpho blue!

rpho Enters Maker’s Endgame. @MakerDAO is deploying $100m DAI to sUSDe/DAI and USDe/DAI markets on Morpho Blue via @sparkfinance's DAI MetaMorpho Vault, providing unmatched liquidity for users who want to increase exposure to @ethena assets. Keep reading for more details.

sUSDe transfer update: ✨ $sUSDe APY 35.2% 💫 Every dollar of value accrued to the protocol from the assets backing the staked USDe were transferred in full to the sUSDe staking contract

Early Access --✨The Modular Application Thesis✨ w/ @PaulFrambot and @delitzer Can the DeFi Money Market sector be disrupted? Can @MorphoLabs compete with incumbents like Aave & Compound? Find out 👇 (for Bankless Citizens only) bankless.cc/ModularAppThesis

i just spent 8 hours writing rebuttals for a set of reviews that were produced in <1 hour of combined effort of 4 reviewers. if i'm lucky, they may spend another 5 minutes reading it, and MAYBE one of them may raise their score by 1 bit. HOW IS THIS A GOOD USE OF ANYONE'S TIME???

I think AI agentic workflows will drive massive AI progress this year — perhaps even more than the next generation of foundation models. This is an important trend, and I urge everyone who works in AI to pay attention to it. Today, we mostly use LLMs in zero-shot mode, prompting a model to generate final output token by token without revising its work. This is akin to asking someone to compose an essay from start to finish, typing straight through with no backspacing allowed, and expecting a high-quality result. Despite the difficulty, LLMs do amazingly well at this task! With an agentic workflow, however, we can ask the LLM to iterate over a document many times. For example, it might take a sequence of steps such as: - Plan an outline. - Decide what, if any, web searches are needed to gather more information. - Write a first draft. - Read over the first draft to spot unjustified arguments or extraneous information. - Revise the draft taking into account any weaknesses spotted. - And so on. This iterative process is critical for most human writers to write good text. With AI, such an iterative workflow yields much better results than writing in a single pass. Devin’s splashy demo recently received a lot of social media buzz. My team has been closely following the evolution of AI that writes code. We analyzed results from a number of research teams, focusing on an algorithm’s ability to do well on the widely used HumanEval coding benchmark. You can see our findings in the diagram below. GPT-3.5 (zero shot) was 48.1% correct. GPT-4 (zero shot) does better at 67.0%. However, the improvement from GPT-3.5 to GPT-4 is dwarfed by incorporating an iterative agent workflow. Indeed, wrapped in an agent loop, GPT-3.5 achieves up to 95.1%. Open source agent tools and the academic literature on agents are proliferating, making this an exciting time but also a confusing one. To help put this work into perspective, I’d like to share a framework for categorizing design patterns for building agents. My team AI Fund is successfully using these patterns in many applications, and I hope you find them useful. - Reflection: The LLM examines its own work to come up with ways to improve it. - Tool use: The LLM is given tools such as web search, code execution, or any other function to help it gather information, take action, or process data. - Planning: The LLM comes up with, and executes, a multistep plan to achieve a goal (for example, writing an outline for an essay, then doing online research, then writing a draft, and so on). - Multi-agent collaboration: More than one AI agent work together, splitting up tasks and discussing and debating ideas, to come up with better solutions than a single agent would. I’ll elaborate on these design patterns and offer suggested readings for each next week. [Original text: deeplearning.ai/the-batch/is…]

2-year Postdoc. Come & work with us & @turinginst on Game Theory (GT), Equilibrium Computation, Multi-agent Reinforcement Learning, Policy Space Response Oracles, and the Gambit Project (software tools for GT, @thegambitproj): turing.ac.uk/research/resear… jobs.ac.uk/job/DGK292/resear…

Great to see @MorphoLabs is paving the way to lending with scalable risk management.

Why trade on margins if you can borrow? medium.com/@simtopia/a76d623… @aave @compoundfinance @MorphoLabs

A must if you want to properly learn modern RL theory.

Great work!

Launching Lean Co-pilot for LLM-human collaboration to write formal mathematical proofs that are 100% accurate. We use LLMs to suggest proof tactics in Lean and also allow humans to intervene and modify in a seamless manner. github.com/lean-dojo/LeanCop… Automating theorem proving remains a hard challenge. Current LLMs are not reliable for math and reasoning tasks since they make mistakes and suffer from hallucinations. Hence, until now, mathematical proofs have been mostly manually derived and require careful verification. Theorem provers like Lean can formally verify each step of the proof but are laborious for humans to write in Lean. Using LLMs to automate suggestions of Lean proof tactics speeds up proof synthesis significantly, and we can incorporate human inputs only when needed through our Lean Co-pilot. - LLMs can suggest proof steps, search for proofs, and select useful lemmas from a large mathematical library. - Lean Copilot is easy to set up as a Lean package and works seamlessly within Lean’s VS Code workflow. - You can use our built-in models from LeanDojo or bring your own models that run either locally (w/ or w/o GPUs) or on the cloud. leandojo.org/ - Models & tools open-sourced under the MIT license Lean Copilot makes LLMs more accessible to Lean users, which we hope will initiate a positive feedback loop where proof automation leads to better data and ultimately improves LLMs on mathematics. @KaiyuYang4 @p_song1 @Caltech

New research from Simtopia team! papers.ssrn.com/sol3/papers.… We study the basic economics underlying the utilization-based interest rates in decentralised lending and how the interest rate model can be updated to give a desired level of utilisation.

Andrew Ng: “Large tech companies would rather not have to try to compete with open source, so they're creating fear of AI leading to human extinction….It's been a weapon for lobbyists to argue for legislation that would be very damaging to the open-source community”

Really hard to overstate how ridiculously excited I am about the next few months for @vegaprotocol. For those that don't know, Vega is the OG DeFi appchain. Yep, what market leaders like @dYdX recently realized, Vega already knew 3 years ago. - Vega's vision has always been to allow anyone, anywhere, to create and participate in markets. Best part? It's all completely permissionless, and very efficient. - Vega's token design is by far the strongest I've seen in the space. Token captures governance, network revenue, and consensus - holders become real owners of the network. - Have shipped like crazy since September, and the next few months will be full of important milestones for the network. - Are working on what I believe to be some of the most exciting product improvements. - Hands down one of the strongest teams in this industry. Really. I will be posting more about Vega for the next few months and closely follow their progress on here. I'm passionate about Vega and about the people building it.