2026
an archive of posts from this year
| Oct 11, 2026 | MemoryTax: The Bitter Lesson of Agent Memory |
|---|---|
| Oct 06, 2026 | Qwen3.8-27B on One RTX 4090 |
| Jul 23, 2026 | Make Sand Think |
| Jul 15, 2026 | 我实现了 GPT-5.6-sol Ultra 自由 |
| Apr 30, 2026 | LoRA as Parametric Memory: a failed experiment |
| Apr 14, 2026 | AI as Research Copilot |
| Mar 01, 2026 | Where Is Linear Attention Actually Slow on GPUs? |
| Mar 01, 2026 | Linear Attention 在 GPU 上到底慢在哪 |