Mathy AI Substack

Mathy AI Substack

Home
Notes
Archive
About
Finetuning is a Memory Wipe. This Is How You Stop It
Scaling Laws for Forgetting When Fine-Tuning Large Language Models, Mike’s Daily Paper: 19.08.25
 • 
Mike Erlihson, Mathy AI
The End of Transformer Babysitting: Forging Stability Without the Hacks.
Mike’s Daily Paper: 19.08.25 - A New Foundation for Stable Transformers: Enforcing Lipschitz Bounds
 • 
Mike Erlihson, Mathy AI
When an LLM Stops Talking and Starts Deleting Files — Enter the Age of LAMs
Omri & Mike’s Daily Paper: 16.08.25, Large Action Models: From Inception to Implementation
 • 
Mike Erlihson, Mathy AI
 and 
Omri Sapir
The Rise of the AI-Augmented Generalist: Navigating the Deep Tech Landscape
The future won't be about who knows the most about one thing, but who can leverage AI to connect everything. But specialists will continue to be vital…
 • 
Mike Erlihson, Mathy AI
The Reasoning Illusion: FormulaOne Exposes the Algorithmic Blind Spot of LLMs
Mike's Daily Paper: 14.08.25, FormulaOne: Measuring the Depth of Algorithmic Reasoning Beyond Competitive Programming
 • 
Mike Erlihson, Mathy AI
We've Been Aligning LLMs All Wrong. The Solution is Deceptively Simple
Mike's Daily Paper: 13.08.25, Checklists Are Better Than Reward Models For Aligning Language Model
 • 
Mike Erlihson, Mathy AI
Beyond Quadratic: A Deep Dive into the Landscape of Efficient Attention
Mike’s Daily Paper: 08.08.25: Efficient Attention Mechanisms for Large Language Models: A Survey
 • 
Mike Erlihson, Mathy AI
The Glass Wall of Knowledge: Why Today's AI Can't Make True Discoveries
We haven't built (yet) an artificial scientist; we've just built a very expensive mirror that only reflects what we already know.
 • 
Mike Erlihson, Mathy AI
It’s Not Just What You Prompt, It’s Where
Mike’s Daily Paper: 06.08.25 - Where to show Demos in Your Prompt: A Positional Bias of In-Context Learning
 • 
Mike Erlihson, Mathy AI
Physics Meets AI: How a New Model Learns Language Without Predicting a Single Token.
Mike’s Daily Paper: 04.08.25Rethinking Transformers Through the Lens of Physics: The Rise of Energy-Based Models
 • 
Mike Erlihson, Mathy AI
Do All Tokens Need the Same Amount of "Thinking"? Mixture-of-Recursions Says No.
Mike's Daily Paper: 02.08.25 - Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation
 • 
Mike Erlihson, Mathy AI
Thinking Without Words: The Architectural Revolution(or not) AI Has Been Waiting For
Mike's Daily Paper: 01.08.25 - Hierarchical Reasoning Model
 • 
Mike Erlihson, Mathy AI
© 2025 Mike E.
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture