After playing with fine-tuning gpt-2 for some simple classification tasks, my honest take is that for most *supervised* ML projects you really, really, really don’t need the LLMs. For simple image/text/signal problems use a pre-trained nn, such as one of the deep NNs trained on
RESEARCH
-
Tesla Optimus Robot Trained in Simulation Transfers to Real World Zero-Shot
By
–
Optimus n’a jamais été entraîné dans le monde réel.
— VISION IA (@vision_ia) 5 juin 2025
Tous ses mouvements viennent de la matrice, un monde équivalent à la simulation.
Puis transférés en zero-shot dans le robot physique.
Bienvenue dans l’ère où les machines apprennent dans leur propre matrice.
Et ça… https://t.co/EsJk26XT5y pic.twitter.com/hsDs4vVOvJOptimus n’a jamais été entraîné dans le monde réel. Tous ses mouvements viennent de la matrice, un monde équivalent à la simulation. Puis transférés en zero-shot dans le robot physique. Bienvenue dans l’ère où les machines apprennent dans leur propre matrice. Et ça
-
Personalization and Student Enjoyment Transform Learning Outcomes
By
–
My takeaways:
– Personalization can have a huge impact
– We shouldn’t underweight the importance of helping students ENJOY learning
– Not all subjects are created equal for personalization! -
Textbooks versus personalized learning in government education
By
–
There was one subject where textbook still reigned supreme: government. The paper has an interesting hypothesis that the subject was just the least broad/hardest to tie to personalized interests like Kpop or gaming
-
Podcasts outperform textbooks for student learning engagement
By
–
Students who got to listen to a podcast, personalized or otherwise, also enjoyed it way more than the textbook group. One quote that really stuck with me
-
Personalized AI Podcasts Boost Student Comprehension in Education
By
–
For philosophy and psych, the group that got personalized podcasts saw it pay off – they scored higher on comprehension quizzes than students who learned from either the non-personalized podcasts or the textbook.
-

AI Personalized Learning: Transforming Education for Every Student
By
–
If you’re lucky, you’ve come across a teacher or mentor in your life who just *got* how you learn. What if AI could do that – for everyone? Research is rolling in on what a difference that could make for students. Take this paper from the CHI conference. 1/
-
AI-Generated Podcasts Beat Textbooks for Student Learning
By
–
Researchers took 180 US college students + divided them up:
– 60 read a textbook chapter (on philosophy, psych, or gov)
– 60 listened to an AI-gen podcast about the chapter
– 60 listened to a personalized AI-gen podcast based on (self-reported) learning style/interests/etc. -
Pretraining: The Beautiful Art of Learning to Compress Knowledge
By
–
pretraining, the act of learning-to-compress the entirety of human knowledge and thereby creating a general-purpose model that can simulate any natural process, is really a beautiful learning paradigm. everything is just going to get messier & more complicated from here
-
Live Webinar: Interpreting LLMs with DLBacktrace Technology
By
–
🔴 We're Live! Join our webinar: Inside the Black Box – Interpreting LLMs with DLBacktrace (DLB) by AryaXAI 🧠 Model reasoning beyond attention 📊 Benchmarking insights 🎙️ Live Q&A Jump in: 🔗 Zoom Link: hubs.la/Q03qTnK_0 #AI #LLMs #Explainability #DLBacktrace