Loka’s Post

🚀 Excited to share our latest blog post on "Simple and Effective Transfer Learning for Neuro-Symbolic Integration"! Dive into the world of Neuro-Symbolic Integration and discover how our proposed transfer learning method addresses key challenges in generalization and reasoning tasks. Check out the full paper here: https://bit.ly/49HyKxu #NeuroSymbolicIntegration #TransferLearning #DeepLearning

Are you familiar with Hugging Face Leaderboard? They take on the daunting task of measure and rank #opensource #LLMs from capability and accuracy space across many factors including: * Multitask Language Understanding * PhD-level in biology, physics, chemistry, etc. * Reasoning and very long-range context parsing * MATH high-school-level competition problems * Capability to follow formatting instructions * Multistep arithmetic and algorithmic reasoning Top 3 Winners: RankNew Leaderboard Ranking 1. ⭐Qwen/Qwen2-72B-Instruct2 2. meta-llama/Meta-Llama-3-70B-Instruct3 3. microsoft/Phi-3-medium-4k-instruct Intrigued yet, you can learn more on what these LLMs have endured through in the link below: https://lnkd.in/e2KKRXZ5

Reinforcement Learning is cool, but can you do it on environments that require you to act in microseconds? This is the topic of my talk at the Reinforcement Learning for Autonomous Accelerators (RL4AA) collaboration workshop that took place in February in the wonderful Salzburg. Thanks to the organizers and all participants for the great opportunity! Enjoy, and let me know what you think! https://lnkd.in/dMNcUViX

Happy Leap Day! My lates blog post, released Monday, features a recent webinar on "Strategies and Tools for Making eLearning Part of Your Publishing Program", with presentations by Paul G. from the American Medical Association, Elaine Tinkelman from the American Chemical Society, and Kelly Lake, Chief Strategy Office from EI. Also included, a mainstream podcast that includes interviews with Brian Nosek, from the Center for Open Science, and Ivan Oransky, of Retraction Watch, discussing ways to combat academic fraud. There is a recently published report from Ithaka S R on digital transformation of scholarly publishing. There are a couple of articles on open science; one reporting on the state of open peer review at PLOS, and the other an OSF preprint reporting on the takeover of open access by commercial interests. There is an interesting new "Rxiv" for storing and sharing search strategies, developed by CABI Digital Library, that use PIDs in a cool way. And of course, there is also a long list of AI articles that I recently read, which I've grouped under the header AYE YAI YAI. https://lnkd.in/eTn9S8wy

Excited to share my recent research paper on "Theoretical Framework for Cloudburst Prediction," conducted under the mentorship of Anima Sharma, which was presented at the 4th International Conference. 🌦️🔍 Our study explores innovative methods to predict cloudbursts using advanced theoretical frameworks. Grateful to Anima Sharma for her invaluable guidance throughout this journey. #ResearchPaper #CloudburstPrediction #InternationalConference #Gratitude #ScienceAndTechnology

Successful Scientist Circles, Video 1: Routine/Process 🎥 🗣️🎙️ A #ScientistCircle is a great way to get students involved in discussion on equal terms. It creates a Circle of Trust in the classroom.  In this edition of #PedagogyOnDeck Brian Klaft talks about the process that makes a Scientist Circle a successful discussion tool. #CircleOfTrust ⭕️😄

In the proposal you describe how you're going to change the world, solve great problems and win the Nobel Prize. In reality you make a very small, but still important, contribution to the body of knowledge. #PhDchat #ECRchat #postdoc

As I always say, Reinforcement Learning! These are the methods in machine learning to which I have been dedicated and to which I'm passionate.

🚀 Exciting Learning Journey with MIT OpenCourseWare! 🚀 I’m thrilled to share that I’ve started the MIT OpenCourseWare course on Algorithmic Thinking and Peak Finding. 📚✨ In the first lecture, we dive into the fascinating world of algorithms, exploring how to efficiently solve the Peak Finding problem. This involves identifying a peak element in a one-dimensional array, where each element is greater than or equal to its neighbors. The lecture covers both linear and binary search approaches, highlighting their complexities and efficiencies. Key takeaways: Algorithmic Thinking: Developing a structured approach to problem-solving. Peak Finding: Understanding and implementing efficient algorithms. Complexity Analysis: Comparing the performance of different algorithms. This course is a fantastic opportunity to deepen my understanding of algorithms and enhance my problem-solving skills. Looking forward to applying these concepts in real-world scenarios! 💡🔍 https://lnkd.in/dYY6pQWM #MITOpenCourseWare #Algorithms #PeakFinding #ContinuousLearning #TechEnthusiast #ProblemSolving

🌍In a world obsessed with productivity charts, Learning Labb Research Institute decided to hack happiness today. We unleashed a secret weapon : tiny notes with ridiculously cute compliments. The result? An explosion of sunshine smiles 🌤🥰that could rival a supernova.  🎥This candid video isn't just about spreading joy; it's a scientific experiment proving the power of positivity as the ultimate learning tool. Get ready for a contagious dose of dopamine as we celebrate the adorable humans who make Learning Labb Research Institute the coolest place to, well, learn! 😃💫 #Positivevibe #Smiles #Zell #Positivity #cutenote #compliment #Learninglabbresearchinstitute #Llri #LLRI #cutesmiles #littlethingsmatter #littlething #ScienceOfSmiles