INDEX

Explanations

the word "in" and its variations within a context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 purpoſe

-0.66

 pleaſure

-0.64

MutableLiveData

-0.60

 itſelf

-0.59

 himſelf

-0.58

 myſelf

-0.58

 chofe

-0.57

 ſta

-0.57

 faſt

-0.56

 raiſ

-0.55

POSITIVE LOGITS

in

1.36

In

1.11

In

0.98

IN

0.94

in

0.91

 isIn

0.83

isIn

0.81

 في

0.79

ใน

0.76

 σε

0.73

Activations Density 1.047%

the word "in" and its variations within a context

No Comments

No Known Activations