INDEX

Explanations

repeated instances of the word "of"

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

a lot of

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 TType

-0.48

FieldBuilder

-0.45

 chrétien

-0.45

CommandText

-0.45

awtextra

-0.45

RegressionTest

-0.44

Predecesor

-0.44

pushFollow

-0.43

EndContext

-0.43

adaptiveStyles

-0.42

POSITIVE LOGITS

 many

0.75

 Many

0.70

of

0.70

Many

0.70

 MANY

0.69

 많은

0.69

 nice

0.65

many

0.63

 things

0.63

Banyak

0.60

Activations Density 0.027%

repeated instances of the word "of"

a lot of

No Comments

No Known Activations