AI Research Posts

FOD#61: AI Fall: Time to Build

Let's get even more practical

14+ Text-to-audio models

A curated list of 14+ text-to-audio AI models: TTS (VALL-E, WaveNet, Bark), music generation (MusicLM, Suno, Jukebox), and sound effects models explained.

Guest Post: Optimizing RAG Applications: Methodologies, Metrics, and Evaluation Tools*

How to optimize RAG: document parsing, chunking, embedding models, retrieval, and evaluation metrics — RAGAS, TruLens-Eval, LlamaIndex, and Phoenix.

9 research papers about synthetic data generation in AI

9 research papers on synthetic data for LLM training: surveys on privacy-preserving generation, Nemotron, DataDreamer, and more.

AI and Search Engines: Petr Ermakov on YandexGPT and the Future of Search

How Yandex built YandexGPT: training on 20TB of data, YaFSDP for GPU efficiency, and how AI is reshaping search. Interview with Petr Ermakov, Yandex

What is LongRAG framework?

LongRAG uses 4K-token retrieval units instead of 100-word chunks, reducing corpus size 30×. How LongRAG architecture works and how it compares to standard RAG.

FOD#59: The Art of Crafting AI with Synthetic Data

plus a collection of thought-provoking essays, important research papers, and news from usual suspects

10+ Research papers to learn more about Vision Language Models (VLMs)

What is KAN?

KAN (Kolmogorov-Arnold Networks) replaces fixed activation functions with learnable splines. How KAN works, how it compares to MLP, and where it falls short.

FOD#58: Are multimodal systems the future of AI?

plus AI-influenced pivots, cool news from the usual suspects and the best weekly curated list of research papers

20+ Large Language Models (LLMs) for Finance