INDEX

Explanations

occurrences of the word "to."

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Autoritní

-0.59

stateMutability

-0.54

PropertyChanging

-0.53

Jeografia

-0.53

autogui

-0.52

usermodel

-0.50

 estekak

-0.50

InSection

-0.49

 сельского

-0.49

 policiales

-0.47

POSITIVE LOGITS

 للمعارف

0.53

 Explo

0.51

 toată

0.49

 OnInit

0.47

 begrü

0.47

 seeing

0.43

 fulfill

0.43

 Gelegenheit

0.42

 zobac

0.41

 neler

0.41

Activations Density 0.004%

occurrences of the word "to."

No Comments

No Known Activations