| ▲ | ijustlovemath an hour ago | |
Why people don't treat categorizing model activations as a science is beyond me. | ||
| ▲ | willy_k 12 minutes ago | parent | next [-] | |
Mechanistic interpretability is a field that studies exactly this. | ||
| ▲ | david-gpu 21 minutes ago | parent | prev [-] | |
Please, go on. How would you do it, specifically? | ||