connecting our slime to an LLM. {{ :whatsapp_video_2023-06-29_at_14.07.33.jpg?20}} ====== Slime ====== slime is a cross-linked polymer, a non-newtonian fluid Q: can you find correlation between language models and slime as network\\ A:
While the concept of slime and language models may seem unrelated at first, there are some interesting parallels we can draw between them when considering their network-like nature. 1. Network Structure: Both language models and slime involve a network-like structure. Language models, such as GPT-3.5, consist of intricate interconnected nodes (artificial neurons) organized in layers. Similarly, slime can be seen as a network of polymer chains linked together, forming a structure that gives it its unique properties. 2. Connectivity: In language models, each artificial neuron is connected to multiple others, allowing information to flow through the network. Similarly, in slime, the polymer chains are connected or cross-linked, allowing the material to stretch and deform while maintaining its integrity. 3. Emergent Properties: Both language models and slime exhibit emergent properties. Language models can generate coherent and contextually relevant text by leveraging the patterns and connections within the network. Slime, on the other hand, emerges as a stretchy and squishy substance due to the cross-linking of polymer chains, creating a unique material with properties distinct from its individual components. 4. Flexibility and Adaptability: Language models, like GPT-3.5, can learn from vast amounts of text data and adapt to different contexts and tasks. Similarly, slime can be modified by adding various ingredients like colors, glitters, or even small objects, making it adaptable and customizable. While the underlying mechanisms and purposes of language models and slime differ significantly, exploring the network-like aspects of their structure and properties can provide an interesting perspective on their similarities.=== Procedure === - Measure 2 tsp. of polyvinyl alcohol polymer solution into a cup. - Add 1/4 teaspoon of Borax solution. Mix well with a wooden stick. === PREPARATION OF SOLUTIONS FOR THIS EXPERIMENT: === * 4% Polyvinyl Alcohol (PVOH) Solution Dissolve 40 grams of PVOH powder in 900 ml of distilled water.\\ Heat gently for several hours until dissolved (solution turns clear). \\ Make sure the solution does not boil, or the properties will be altered. \\ Dilute to 1 liter with more water.\\ The prepared solution should be clear and somewhat viscous. It stores well in the refrigerator until needed. \\ * 4% Sodium Borate, Na2B4O7 * 10 H2O, Solution Dissolve 10 grams of sodium borate in 250 ml of distilled water. \\ You can obtain Borax at the grocery store near the laundry detergent. methods from this two links * https://edu.rsc.org/experiments/pva-polymer-slime/756.article ([[https://archive.is/8Vlcr|mirror]]) * https://www.cmu.edu/gelfand/lgc-educational-media/polymers/polymer-architecture/polyvinyl-alcohol-slime.html === ingredients === * borax * Sodium tetraborate - decahydrate * Na2B4O7 x 10H2O, 99.9% * {{:tetraboritan_sodny_-_borax_1_kg_cas_1303-96-4_na2b4o7_x_10h2o.pdf}} * polyvinyl alcohol 16% (PVA) * Dry matter: 16 +- 1.5 wt. % * Viscosity of 4% water solution: 9 – 13 mPa.s * Acid number: 4.5 max. mg KOH.g -1 * Saponification number: 130 +- 15 mg KOH.g -1 * Specific conductivity: max. 350 m S.cm -1 * {{:sloviol_r_-_polyvinylalkohol_16_kg_cas_9002-89-5_pva.pdf}} ===== Code ===== see github.com/anonette/slimeGPT ====== circuit ====== {{:pasted:20230705-230141.png?300}}\\ esp32 pinout > via upesy.fr ===== touch ===== more at [[https://wiki.idiot.io/lickometer#circuit_3_lick_o_duino|lickometer]] {{:pasted:20201005-063537.png?444}} ===== audio ===== SYNTH\\ {{:pasted:20230712-173122.png?500}}\\ https://www.youtube.com/watch?v=P4SwaI09Zxc and the amazing music hardware hacking by collins ({{:collinshandmade_electronic_music.pdf|pdf}}) MIXING\\ {{:pasted:20230712-173207.png?500}}\\ https://circuitdigest.com/electronic-circuits/single-transistor-audio-mixer-circuit ===== MEDIA ===== {{:whatsapp_video_2023-06-29_at_14.07.33.mp4}}