Skip to content
mini-lab
Features of prelude-2

Feature #1135

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

Fires on
0.25% of tokens
Mostly in
story
Layer
after block 4

Where it fires

  • story 77%
  • agent 11%
  • chat 10%
  • python 2%
  • worksheet 1%

Makes the model say

was they she he day couldat if

Makes it not say

lenpatternnder reverFixedllo E upon

Strongest examples

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

  1. story 6.1
    , there was a little girl named Amy. Amy had a toy rod that she loved to play
  2. story 5.9
    boy named Tim. He had a toy ghost that he loved to play
  3. story 5.81
    there was a little boy named Tim. He had a black toy car that he loved very much
  4. story 5.74
    there lived a little boy named Tim. Tim had a toy box that he loved to fill
  5. story 5.67
    time, there was a boy named Tom. He had a toy gun that could shoot soft balls
  6. chat 5.52
    was a big, friendly bear named Bobo. Bobo had a toy car that he loved to play
  7. story 5.46
    a time, there was a boy named Tim. Tim had a pencil that he loved very much
  8. story 5.38
    there was a girl named Lucy. She had a pretty uniform that she wore to school
  9. story 5.37
    was a gifted little girl named Amy. Amy had a pink dress that she loved very much
  10. story 5.37
    , there was a small boy named Tim. Tim had a penny that he loved very much
  11. story 5.34
    , there was a little girl named Mia. Mia had a yellow coat that she loved very much
  12. story 5.27
    time, there was a little girl named Lily. She had a doll that she loved so much