INDEX

Explanations

references to the concept of "everything."

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

بوابة

-0.66

 لئے

-0.65

 אף

-0.64

filterBy

-0.64

vox

-0.63

Hollow

-0.63

ȡ

-0.63

βο

-0.61

amerikanischer

-0.61

oucí

-0.61

POSITIVE LOGITS

 everything

2.44

everything

2.38

Everything

2.37

 Everything

2.36

 EVERYTHING

2.06

THING

1.68

 tudo

1.55

 things

1.49

Alles

1.49

 alles

1.44

Activations Density 0.031%

references to the concept of "everything."

No Comments

No Known Activations

references to the concept of "everything."

No Comments

No Known Activations