# Jan 5, 2025

## Daily

### The day I taught AI to read code like a Senior Developer

- **AI's code analysis improved dramatically** by adopting a **context-aware grouping system**, enabling it to analyze code like a senior developer rather than a novice, leading to insights about potential issues that were previously overlooked.

### No more needles Tracking blood sugar on your wrist

- **University of Waterloo engineers** have developed a **non-invasive glucose monitoring system** that utilizes miniaturized radar technology, allowing for continuous tracking of blood sugar levels without the need for needles or skin penetration.

### Human study on AI spear phishing campaigns

- **AI spear-phishing campaigns** achieved a **click-through rate exceeding 50%**, demonstrating their effectiveness compared to traditional methods, with AI-generated emails performing on par with human experts and significantly better than control groups.

### [R] High-performance deep spiking neural networks with 0.3 spikes per neuron

- **Deep spiking neural networks** can achieve high-performance classification with **less than 0.3 spikes per neuron**, demonstrating energy efficiency comparable to biological brains while maintaining accuracy on datasets like MNIST and CIFAR.

### [P] Noteworthy AI Research Papers of 2024 (Part One)

- **Six influential AI papers** from January to June 2024 highlight advancements in LLMs, including the **Mixtral 8x7B**, a Sparse Mixture of Experts model that outperforms Llama 2 70B and GPT-3.5 across benchmarks, showcasing the potential of MoE architectures for efficient scaling.

### [R] How Barlow Twins avoid embeddings that differ by affine transformation?

- **Barlow Twins (BT) loss** minimizes when the **cross-correlation matrix** equals the **identity matrix**, ensuring that embeddings from different augmentations remain invariant to affine transformations, thus promoting robust representation learning.

### Hornet, RNA structure visualization method correlates sequence and 3D topology

- **HORNET** is a novel RNA structure visualization method that utilizes **atomic force microscopy (AFM)** and **deep neural networks** to accurately characterize the 3D topological structures of large, flexible RNA molecules, overcoming limitations of traditional methods.

### Waymo Robotaxis Much Safer Than Any Human-Driven Cars

- **Waymo's robotaxis demonstrate an astonishing** **88% reduction in property damage claims** and **92% reduction in bodily injury claims** compared to human drivers, showcasing their superior safety performance.

### [Project] Finding inputs where deep learning models fail

- **Deep learning models struggle** with distinguishing between dogs and cats in the CIFAR-10 dataset, revealing significant limitations in modern architectures when compared to human expertise.