bycloud

The Chinese AI Iceberg

Исследователи получают действительно креативное обучение LLM [Прогнозирование порядка токенов]

Искусство эффективного обслуживания студентов магистратуры права: объяснение контекстной инженерии

Возможно ли вообще провести обратную разработку данных обучения ИИ?

New AI Paradigm?! Energy-Based Transformers Explained

New AI Meta: Train LLMs To Explore On "Hard" Tokens [RLVR + Entropy]

Kimi K2 Technical Breakdown: How It Challenged AI’s 7-Year Status Quo

a rather unusual announcement

If AI's So Good, Why Can't It Make Its Own Data? [SEAL]

10x Faster Than Standard LLM!? DiffusionLM Explained

The Side Effects of Overusing ChatGPT For Homework

POV: Chinese AI Lab Teaching Everyone How To Save Millions of Dollars

What If We Remove Tokenization In LLMs?

When AI Is Designed Like A Biological Brain

The LLM's RL Revelation We Didn't See Coming

1-Bit LLM: The Most Efficient LLM Possible?

The REAL AI Architecture That Unifies Vision & Language

How DeepSeek Built The Current "Best" Math Prover AI

AI News: DeepSeek-R1 V2, the new open source SoTA!

Why would anyone let LLMs predict 4 tokens at once? Multi-Token Prediction Explained

The Biggest "Lie" in AI? LLM doesn't think step-by-step

Attention Sink: The Fluke That Made LLMs Actually Usable

Anthropic Just Proved Reasoning AIs Would Silently Cheat

how to deal with reasoning AIs that just can't stop yapping

Meta AI did something WILD again... wtf is Next Concept Prediction?

I built a chatbot that can answer all the new AI concepts

Gemini 2.5 Pro is just the best choice for AI right now

OpenAI's Cancels GPT-4.5!? First Look At GPT-4.1 with 1M Context

Did OpenAI Accidentally Make Vibe Arting Real...?

Meta's new Llama-4 is Full of Controversies...