Arpit Bhayani’s Post

The devil is in the details. And details are not always generated. Use your tools wisely

while summaries are a great way to get the crux of the subject, we cannot learn a lot about the topic from the crux, so it's very important to keep a tab on learning instead of consuming the content

Today?Why Study Critical Theory Today? Comprehensive Final Assignment: Prompt En

Can LLMs make us critical thinkers? TreeInstruct reorients LLMs to be instructors that guide students socratically to solve problems, instead of assistants that provide direct answers. Paper: https://lnkd.in/dG48E3Zm Code: https://lnkd.in/dp3KQ9h6 We apply TreeInstruct to code debugging. Prior works directly give away bugs/fixes, assume single-turn conversations, and only work for one bug. We create a realistic, multi-bug dataset, where the bugs are mutually dependent. TreeInstruct estimates the knowledge a student needs to debug their code and devises a conversation plan. It then dynamically constructs a question tree based on its interactions with the student, navigating the knowledge state space till the student comprehends & fixes all bugs. TreeInstruct is preferred 78.43% of the time. It solves 14.09% more bugs across all settings, and our questions are 14.18% better at addressing bugs, maintaining relevance, and ensuring logical conversation flow. TreeInstruct also adapts to human students of varying backgrounds. Collaboration with: Ishika Agarwal, Dilek Hakkani-Tur, Jiawei Han Special Thanks: Mihir Kavishwar, Krish Agarwal, Sonia Agarwal

Can LLMs make us critical thinkers? TreeInstruct reorients LLMs to be instructors that guide students socratically to solve problems, instead of assistants that provide direct answers. Paper:  https://lnkd.in/gYgbVQJG Code: https://lnkd.in/gqSmsJjA We apply TreeInstruct to code debugging. Prior works directly give away bugs/fixes, assume single-turn conversations, and only work for one bug. We create a realistic, multi-bug dataset, where the bugs are mutually dependent. TreeInstruct estimates the knowledge a student needs to debug their code and devises a conversation plan. It then dynamically constructs a question tree based on its interactions with the student, navigating the knowledge state space till the student comprehends & fixes all bugs. TreeInstruct is preferred 78.43% of the time. It solves 14.09% more bugs across all settings, and our questions are 14.18% better at addressing bugs, maintaining relevance, and ensuring logical conversation flow. TreeInstruct also adapts to human students of varying backgrounds. Collaboration with: Priyanka Kargupta, Dilek Hakkani-Tur, Jiawei Han Special Thanks: Mihir Kavishwar, Krish Agarwal, Sonia Agarwal

This article explores two fundamental frameworks that can fast-track your journey to becoming an expert: The 4 Stages of Competence and Bloom’s Taxonomy. The 4 Stages of Competence framework outlines the psychological phases one experiences while learning a new skill, ranging from unawareness to proficiency. Conversely, Bloom's Taxonomy offers a tiered model for educational goals, leading learners through levels from simple memorization to advanced analysis and synthesis. https://lnkd.in/gKNC4p_w

Day 37: The Power of Revisions 📚 Today was all about revisiting and reinforcing the concepts I've covered so far in my DSA journey. There's something profoundly valuable about taking the time to go back and re-examine what you've learned: 🧠 Why Revisions Matter: Solidifying Understanding: Revisiting old concepts helps to ensure that knowledge is deeply ingrained, not just on the surface. Connecting the Dots: As I revisited topics like Binary Search, Matrix Manipulations, and Array Problems, I noticed how interconnected these concepts are. Seeing the bigger picture is crucial for problem-solving. Confidence Building: Every time I revisit a concept and solve a problem more efficiently than before, my confidence grows. It’s a reminder that learning is a continuous process, and mastery comes with time. 🔍 Areas of Focus: Binary Search Variants: Reinforcing concepts like lower and upper bounds, and understanding their application in various scenarios. Matrix Operations: Going over Set Matrix Zeros, Rotate Image by 90 degrees, and the Spiral Matrix—problems that demand a solid grasp of 2D arrays. Array Manipulations: Revisiting problems like finding the majority element, moving zeros to the end, and understanding permutations. 🔗 The Big Picture: Revisions are not just about repetition—they’re about deepening your understanding and connecting the dots between different concepts. Every problem I solve now feels a little more intuitive, a little less daunting, thanks to the work I've put in. Here's to making the most of the journey and continuing to build a strong foundation! 🚀 #DSA #CodingJourney #Revisions #LearningProcess #TechLife #ProblemSolving #GrowthMindset

With all the resources at our fingertips to learn or develop a new skillset, are we forgetting a lost art? Finding videos, websites and archives of information is so valuable in our careers. It allows us to not only learn more but gives us the tools to make forward strides in our engineering or PM journey. However, don’t count out a great book! Reading books can provide the same value, if not more, than the resources we find here online. Reading helps us understand hard concepts and can indirectly help with our communication and writing skillset. It gives us a moment to pause, think critically and to get familiar with new frameworks or problem solving. All in all, it allows for deeper understanding. Here are two of the best book recs I’ve received this year that I’m currently reading: ✨Deep Work by Cal Newport ✨ The Pragmatic Programmer by Andrew Hunt and David Thomas What are some of your favorite industry books that are a must-read? Share in the comments below! #softwareengineering #resourcesharing #techindustry

🤔💬 DYK we remember things better when we create our own version of the material. 🔄 The 'Generation Effect' shows that actively manipulating new information helps form relationships between concepts, making them easier to recall when needed. Source: Anne-Laure Le Cunff, Ness Labs #teachers #students #learning #teaching #GenerationEffect

Ever found yourself struggling to retain information after hours of studying? 🤔  Don't worry; you're not alone! But what if I told you there's a powerful theory that can revolutionize the way you approach learning? Enter the Depth of Processing Theory! 🧠 This groundbreaking concept suggests that the depth of processing directly impacts memory retention. In simpler terms: the deeper you delve into information, the better you remember it! But how can you harness this theory to supercharge your learning journey? That's where active reading strategies come into play! By actively engaging with the material through summarizing, analyzing, and synthesizing, you're not just skimming the surface – you're diving deep into the heart of knowledge. Imagine reading a dense textbook chapter. Instead of passively flipping through pages, try summarizing each section in your own words. Challenge yourself to analyze key concepts and connect them to real-life examples. And don't forget to synthesize information by creating mind maps or flashcards to reinforce your understanding. By adopting these active reading techniques, you're not just studying – you're sculpting your memory like a master craftsman! Whether you're preparing for exams, mastering new skills, or staying ahead in your career, the power of active reading is your secret weapon. Share your favorite active reading strategies in the comments below and let's inspire each other to reach new heights of comprehension and retention! 🚀🌟 #reading #readingcomprehension #learningjourney #readingchallenge #learninghub