Feature #2907
Fires on “ a” (not labeled yet: read the examples)
- Fires on
- 0.40% of tokens
- Mostly in
- python
- Layer
- after block 4
Where it fires
- python 36%
- chat 23%
- agent 20%
- story 16%
- worksheet 5%
Makes the model say
defSoprint We us read So
Makes it not say
th<|user_start|> areThank2Good felt waiting
Strongest examples
On text the dictionary never trained on. The underlined token is where it fires most.
-
chat 3.12it in New York<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.11I fix a flat tire?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.1I fix a flat tire?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.09<|user_start|>What does DNA stand for?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.03is it in Berlin?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.03the tallest mountain in the world?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 3.01physics to me.<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
python 3.0) == 1⏎assert sub_one(11) == 10⏎assert
-
chat 2.99is the Eiffel Tower?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I
-
chat 2.99I fix a flat tire?<|user_end|><|assistant_start|>Sorry, I'm a
-
python 2.98max(nums)⏎⏎$ python3 demo.py⏎7⏎5⏎$ python
-
chat 2.97is it in Sydney?<|user_end|><|assistant_start|>Sorry, I'm a tiny model: I