Skip to content
mini-lab
Features of prelude-2

Feature #3629

Fires on “ were” (not labeled yet: read the examples)

Fires on
0.14% of tokens
Mostly in
python
Layer
after block 4

Where it fires

  • python 36%
  • story 26%
  • agent 25%
  • chat 11%
  • worksheet 3%

Makes the model say

⏎ aboutke add spnums O all

Makes it not say

assert she<|bos|>arful>>> he special

Strongest examples

On text the dictionary never trained on. The underlined token is where it fires most.

  1. story 2.65
    37 = 91540<|bos|>Tom and Mia were best friends. They
  2. story 2.62
    can only tell short stories and add numbers.<|assistant_end|><|bos|>Tom and Sam were friends. They liked
  3. story 2.52
    8<|think_end|>The answer is 78.<|assistant_end|><|bos|>Tom and Sam are friends. They like
  4. story 2.48
    I can tell short stories and add numbers.<|assistant_end|><|bos|>Tom and Mia were playing in the park
  5. story 2.39
    numbers. Try asking me for a story!<|assistant_end|><|bos|>Tom and Mia are at the beach with
  6. story 2.38
    The answer is 135.<|assistant_end|><|bos|>Jen and Ben were twins who liked to
  7. story 2.37
    a story, or some numbers to add?<|assistant_end|><|bos|>Sara and Tom are friends. They like
  8. story 2.36
    <|think_end|>The answer is 333.<|assistant_end|><|bos|>Ben and Sam were brothers who liked
  9. story 2.35
    The answer is 1706.<|assistant_end|><|bos|>Tim and Sam are friends. They like
  10. story 2.33
    <|tool_end|><|assistant_start|>Created `greeting.py`.<|assistant_end|><|bos|>Sam and Ben are brothers. They
  11. story 2.33
    Goodbye! Have a nice day.<|assistant_end|><|bos|>Anna and Ben were playing in the snow
  12. story 2.32
    scratch on a laptop by mini-lab.<|assistant_end|><|bos|>Tom and Lily were twins who liked to