Mishthi Sachdeva

Hi, I’m Mishthi. 20. Just a girl in tech figuring things out. Into AI, random experiments, and understanding how things work behind the scenes. Learning. Building. Growing. If you’re exploring tech and growing along the way, let’s connect 🤝 @X show this to curious builders🚀

@cat3lyst heyyy let's connect!!

If you're in tech. Let's connect

@AshwinN14729359 heyy let's connect!

If you're in tech, Let's connect🤝

@victoria_echezu heyy let's connect!!

if you are in Tech let's connect and engage 📈🎀

@won__sikkk Both matter, but per request cost usually dominates because inference runs millions of times. That said, techniques like caching, batching, and better request design can reduce how often requests happen, which also saves a lot of compute.

@SachdevaMi94386 Which matters more in practice now, lowering per request cost or reducing how often requests need to happen?

Decoding Tech #14 Training an AI model is expensive. But inference is what runs constantly. Every prompt triggers a forward pass through billions of parameters. Training happens once. Inference happens millions of times. That’s where most real-world compute gets used.

Decoding Tech #13 What actually happens during inference? When you ask an AI a question, the model isn’t “thinking.” It’s running a forward pass through the network and predicting the next token step by step. No learning. Just computation. #DecodingTech

@LearnWithBhavik sure!!

@SachdevaMi94386 Great breakdown. Token limits explain why context disappears mid-chat. Prompt structure really matters when working with LLMs. Let’s connect.

Decoding Tech #12 AI models have token limits. Your prompt + previous messages + the response must all fit inside that window. If it gets too long, older context gets dropped. That’s why AI sometimes forgets things mid-conversation. #DecodingTech

@iamkarank5 @X sure!

@SachdevaMi94386 @X Let's Connect 🤝

Hey @X algorithm, Looking to connect with people interested in: 🎨 Frontend 🧠 Backend 🤖 GenAI ✨ Full Stack ⚙️ DevOps ✅ DSA 🧠 AI/ML 🌐 Web3 📊 Data Science 💼 Freelancing 🐍 Python 🚀 Startups Let’s learn, grow, and build in public 💪🔥

Decoding Tech #11 AI models don’t read text the way we do. They read tokens. A token can be a word, part of a word, or even punctuation. For example: “unbelievable” → un + believ + able Before AI understands language, language is broken into tokens. #DecodingTech

@XtroUI exactly!

@SachdevaMi94386 Transformers truly revolutionized NLP! 🚀 It’s amazing how parallel processing boosts understanding.

Decoding Tech #10 Before transformers, most language models used RNNs. They processed text one word at a time. Transformers changed that. They process all words simultaneously using attention. Parallel processing + better context understanding. #DecodingTech

Happy Holi everyone! 🌸 Wishing you a day full of colours, happiness, and sweet moments with the people you love.

Decoding Tech #9 Why does AI training use GPUs instead of CPUs? CPUs handle tasks sequentially. GPUs handle thousands of calculations in parallel. Training AI means updating billions of parameters at once. AI progress = parallel computation. #DecodingTech

@hey_shivv Thankyou!

@SachdevaMi94386 Keep going

Decoding Tech #8 Pretraining vs Fine-tuning Pretraining teaches a model how language works. Fine-tuning teaches it how to behave. One builds understanding. The other shapes responses. That’s why the same base model can power very different applications. #DecodingTech

@Osayi80858113 @DaKSH18_ done!

@SachdevaMi94386 @DaKSH18_ Alright, done from my end Ensure to always engage, and i will do the same

If you are in tech Let's connect & follow each other

@adefemi_a30197 true! Let's connect!

@SachdevaMi94386 AI sees shapes, not words. Meaning emerges from distances in high-dimensional space. Let’s connect.

Decoding Tech #7 AI doesn’t understand words. It converts them into vectors. Every word (or token) becomes a list of numbers placed in a high-dimensional space. Words with similar meanings end up closer together in that space. Language → geometry → meaning #DecodingTech

@Osayi80858113 @DaKSH18_ yes yes, let's connect!

@SachdevaMi94386 @DaKSH18_ Let's also connect Are you active ?

@DaKSH18_ sure!

@SachdevaMi94386 Let's Mishti