🤖 Three major AI collectives deeply interpret DLMM mechanism Everything is the same, if it's just for daily use, you don't need to know the underlying technical principles. 🚴‍♀️ Just like the vast majority of people believe that they can ride a bicycle, but when it comes to explaining why a bicycle doesn't fall, the scientific community has spent over 100 years providing different explanations. 📈 DLMM is the same. I have been using it for at least a year and a half before, and through the website front-end, it mainly involves selecting pools, rates, liquidity strategies, and so on. Do not understand the underlying principles, use without error. But recently, if you want to integrate DLMM through a program and automate the increase or decrease of liquidity, it is necessary to explore it carefully. In the face of such technical and fundamental issues, when it comes to in-depth interpretation, the use of various AI models in Deep Research is just right. My current problem is: I need to have a deep understanding of the underlying operating principles of DLMM, especially how to integrate its SDK, choose price ranges, and choose liquidity strategies. So, I chose the following tools: 1 Gemini 2.5 Deep Research 2 OpenAI O3 Pro Deep Research 3 Claude Opus 4 Research Let the three major AI tools explore the same issue in depth and provide their respective reports. Based on three major reports, I have updated many of my outdated knowledge: one ️⃣ Several core concepts of DLMM 1. Position (liquidity position) 2. Bin (Price Bucket) 3. Fee (fee model) 4. Strategy (Market making Strategy Module) two ️⃣ Position Each Position is a liquidity range that you deploy. Optional: 1. Balance position 2. Unbalanced position 3. Unilateral liquidity position This is indeed more flexible and free than UniSwap V3. In terms of actual user experience, it is indeed very considerate of LP. three ️⃣ DLMM's Bin Bin's design has several innovations (actually inspired by Trader Joe's Liquidity Book model) 1. Using the formula x+y=L ⋅ P within a single Bin 2. Zero slippage point for token exchange within a single Bin 3. Only the liquidity of Active Bin can obtain transaction fees 4. Only Active Bin has two types of tokens (other inactive Bin only have unilateral liquidity) four ️⃣ Fee The handling fees here are divided into base fees and dynamic fees. A fixed base fee, combined with a dynamically adjusted fee based on the magnitude of volatility, is used to compensate for the unpredictable losses of LP. five ️⃣ Strategy I have re examined these three strategies and corrected my long-standing biases I always think Spot and Curve are the right paths for market making strategies, while Bid Ask is the wrong path 😳） In fact, it can be said that Bid Ask is a mixture of market making strategy and quantitative trading strategy. What scene and how to use it well depend on your skills. After reading the report, it must be said that @ MeteoraAG has indeed put in a lot of effort and thought into the LP matter.