groundy

models & research

  1. may 17modelsLearning, Fast and Slow: What arXiv 2605.12484 Proposes for LLMs That Adapt Continually
  2. apr 27modelsThere Will Be a Scientific Theory of Deep Learning: What arXiv 2604.21691 Argues and Where It Will Lose
  3. apr 22modelsQwen3.6-27B's Dense Architecture Challenges the MoE-Only Playbook for Flagship-Class Coding Models
  4. mar 23modelsChinese AI Models Compared: DeepSeek, Qwen, Kimi, Doubao, and Ernie
  5. mar 23modelsRunning DeepSeek R1 Locally: Hardware Requirements, Quantization, and Real Throughput
  6. mar 14modelsFish-Speech: The Open-Source TTS Model That's Threatening ElevenLabs
  7. feb 26modelsSynthetic Data Is Eating AI Training
  8. feb 26modelsGoogle's TimesFM: A Foundation Model for Time Series
  9. feb 26modelsGemini 2.0 Pro's 2 Million Token Context: What Can You Actually Do With It?
  10. feb 26modelsDeepSeek V3/R1: How Chinese Engineers Matched GPT-4 for $6 Million
  11. feb 26modelsClaude's Web Search Changes Everything for AI Research
  12. feb 26modelsThe Million-Token Context Window: What Can You Actually Do?
  13. feb 18modelsGemini 3.1 Pro: Google's New Reasoning Model Explained
  14. feb 17modelsKimi Claw: Moonshot AI's Answer to Claude and ChatGPT
  15. feb 17modelsWiFi DensePose: Full-Body Tracking Through Walls Using Your Router
  16. feb 14modelsAI Code Generation Benchmarks 2026: Which Model Actually Writes Better Code?
  17. feb 10modelsThe Best AI Models for OpenClaw in 2026