INDEX

Explanations

phrases indicating choice or alternatives

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

 BrowserModule

-0.54

 FontFamily

-0.49

 Delf

-0.45

roit

-0.43

ayashi

-0.43

立ち

-0.43

﴿

-0.42

 consectetur

-0.42

 progressivement

-0.42

 Encore

-0.42

POSITIVE LOGITS

 мәкал

0.79

ſy

0.73

Попис

0.69

 ſtate

0.68

NSError

0.67

 ſta

0.65

Eſ

0.65

LLocation

0.64

 Administrativna

0.64

 faſt

0.63

Activations Density 0.184%

phrases indicating choice or alternatives

No Comments

No Known Activations