INDEX

Explanations

repeated instances of the word "said" and its variants

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 Staub

-0.68

?>">

-0.62

ladin

-0.62

bibfield

-0.61

Gera

-0.61

bibinfo

-0.61

AppCompat

-0.61

FLO

-0.61

bern

-0.60

)|^{

-0.60

POSITIVE LOGITS

 said

1.95

 Saying

1.78

say

1.78

said

1.76

SAY

1.76

 saying

1.75

saying

1.75

SAY

1.75

Said

1.73

 SAID

1.72

Activations Density 0.086%

repeated instances of the word "said" and its variants

No Comments

No Known Activations