Skip to content
mini-lab
Features of prelude-2

Feature #1512

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

Fires on
0.21% of tokens
Mostly in
python
Layer
after block 4

Where it fires

  • python 44%
  • story 41%
  • chat 8%
  • agent 6%
  • worksheet 1%

Makes the model say

m were whoal with is again best

Makes it not say

```]`. Finopx<|bos|>RThis

Strongest examples

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

  1. story 5.97
    7 = 92725.<|bos|>Lily and Tom are friends. They like to
  2. story 5.88
    <|think_end|>The answer is 78.<|assistant_end|><|bos|>Tom and Sam are friends. They like to
  3. story 5.87
    790<|think_end|>790<|assistant_end|><|bos|>Lily and Ben are friends. They like to
  4. story 5.81
    7 = 91540<|bos|>Tom and Mia were best friends. They lived next
  5. story 5.72
    4, 9])⏎9⏎<|bos|>Tom and Ann are friends. They like to
  6. story 5.6
    only tell short stories and add numbers.<|assistant_end|><|bos|>Tom and Sam were friends. They liked to
  7. story 5.55
    _stats.py⏎all tests passed⏎<|bos|>Tom and Jen are friends. They like to
  8. story 5.55
    answer is 1706.<|assistant_end|><|bos|>Tim and Sam are friends. They like to
  9. story 5.48
    <|think_end|>The answer is 63.<|assistant_end|><|bos|>Anna and Ben are friends. They like to
  10. story 5.47
    very happy to have each other as friends.<|bos|>Tom and Lily are friends. They like to
  11. story 5.47
    95 is 2603.<|bos|>Anna and Ben are friends. They like to
  12. story 5.46
    was happy to have a new friend.<|assistant_end|><|bos|>Tom and Mia are friends. They like to