Feature #2487
Fires on “ Python” (not labeled yet: read the examples)
- Fires on
- 0.24% of tokens
- Mostly in
- agent
- Layer
- after block 4
Where it fires
- agent 58%
- python 22%
- story 13%
- chat 5%
- worksheet 1%
Makes the model say
files project definedcommand laptopitymainlower
Makes it not say
help import count promised you be now big
Strongest examples
On text the dictionary never trained on. The underlined token is where it fires most.
-
agent 3.71only knows small Python projects. I
-
agent 3.69can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.68can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.64can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.62I can list, read, run, create, edit and fix the Python files in this project
-
agent 3.62can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.59can't do that: I'm a tiny model that only knows small Python projects. I
-
python 3.57<|bos|># README.md⏎# Scratch⏎⏎A small Python project.⏎⏎
-
agent 3.57can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.55can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.55can't do that: I'm a tiny model that only knows small Python projects. I
-
agent 3.55can't do that: I'm a tiny model that only knows small Python projects. I