chris-hay
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| I Quantized Qwen3.8-27B. “4-Bit” Doesn’t Mean What You Think |
|
1 | 9 | 3 September 2026 |
| I Trained the TinyStories Model to Do Math |
|
1 | 21 | 27 July 2026 |
| The AI Agent Memory That's Impossible to Lie To |
|
1 | 16 | 13 July 2026 |
| GPT-5.6 Sol, FIFA AI & Wall Street’s AI nerves |
|
1 | 17 | 3 July 2026 |
| The Model Doesn't Unpack Its Memory |
|
1 | 23 | 10 June 2026 |
| I Decoupled Attention from Weights - Gemma 4 26B |
|
1 | 22 | 5 May 2026 |