# Feb 2, 2025

### Chatbot Software Begins to Face Fundamental Limitations

- **Large language models (LLMs) exhibit significant limitations in solving compositional tasks**, as evidenced by their poor performance on complex problems like Einstein's riddle, indicating a fundamental cap on their reasoning abilities.

### RLHF Book

- **Reinforcement Learning from Human Feedback (RLHF)** is a pivotal method for enhancing machine learning systems, integrating insights from economics, philosophy, and optimal control to improve language models.

### 3D scene reconstruction in adverse weather conditions via Gaussian splatting

- **WeatherGS** enhances **3D scene reconstruction** by effectively addressing artifacts from adverse weather, utilizing a novel **dense-to-sparse preprocess strategy** to improve clarity in reconstructed scenes.

### Tulu 3 model performing better than 4o and Deepseek?

- The **Tulu 3 model**, released by the **Allen Institute for AI**, reportedly **outperforms** both **4o** and **DeepSeek** in several benchmarks, showcasing advancements in AI model performance.

### Molecular Fingerprints Are Strong Models for Peptide Function Prediction

- **Molecular fingerprints** outperform complex models like GNNs and transformers in peptide classification, achieving **state-of-the-art results** on 126 datasets, including LRGB, without hyperparameter tuning.

### EXAdam: The Power of Adaptive Cross-Moments

- **EXAdam** is an advanced optimization algorithm that enhances the **Adam optimizer** with new debiasing terms, a gradient-based acceleration mechanism, and a dynamic step size formula, leading to improved convergence and robustness.

### Researchers combine holograms and AI to create uncrackable optical encryption

- **Researchers have developed an optical encryption system** that utilizes holograms and neural networks, achieving a level of security that traditional encryption methods cannot match, particularly beneficial for sectors like digital currencies and healthcare.

### Addressing Underthinking in LLMs: A Token-Based Strategy to Improve Reasoning Depth

- This research presents a **novel methodology** for analyzing **underthinking** in large language models (LLMs) by employing **token-level output analysis** to track reasoning consistency, revealing that models switch cognitive approaches every **2-3 reasoning steps** on average.