MACHINE LEARNING - AI Dynamics

Representational Geometry Influences Uncertainty in Large Language Models

By

–

08 May 2026 19h08

"Representational Curvature Modulates Behavioral Uncertainty in LLMs" Most interpretability work studies what features live inside LLMs. But this paper studies something deeper, which is the geometry of the whole evolving representation. The key result is that representational

→ View original post on X — @askalphaxiv

8 May 2026

Monetizing machine learning models with XGBoost

By

@tunguz

–

08 May 2026 18h42

Is your XGBoost model making money for you while you sleep? @trainxgb @tabul_ai

→ View original post on X — @tunguz

8 May 2026

Hardware limitations on sparse neural activation in LLMs

By

@hardmaru

–

08 May 2026 18h29

The human brain is incredibly efficient because it only activates the specific neurons needed for a thought. Modern LLMs naturally try to do this too (> 95% of neurons in feedforward layers stay silent for any given word), but our hardware punishes them for it. One of the most

→ View original post on X — @hardmaru

8 May 2026

Optimiser les LLM avec la sparsité adaptée au GPU

By

@sakanaailabs

–

08 May 2026 18h26

How do we make LLMs faster and lighter? Don’t force the GPU to adapt to sparsity. Reshape the sparsity to fit the GPU! ⚡️

Excited to share our new #ICML2026 paper in collaboration with @NVIDIA: "Sparser, Faster, Lighter Transformer Language Models". This work introduces new… pic.twitter.com/ehByWHIh6I
— Sakana AI (@SakanaAILabs) 8 mai 2026

How do we make LLMs faster and lighter? Don’t force the GPU to adapt to sparsity. Reshape the sparsity to fit the GPU! Excited to share our new #ICML2026 paper in collaboration with @NVIDIA
: "Sparser, Faster, Lighter Transformer Language Models". This work introduces new

→ View original post on X — @sakanaailabs

8 May 2026