INDEX

Explanations

terms indicating significance or importance, especially in a quantitative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

aculture

-0.56

__*/

-0.56

Gambas

-0.55

ex

-0.54

 Forster

-0.53

 tele

-0.53

zam

-0.51

 couple

-0.51

'.$

-0.51

hdashline

-0.50

POSITIVE LOGITS

major

1.66

Major

1.60

 Major

1.52

 major

1.36

 MAJOR

1.35

Minor

1.34

Maj

1.32

MAJOR

1.31

minor

1.29

 Minor

1.24

Activations Density 0.113%

terms indicating significance or importance, especially in a quantitative context

No Comments

No Known Activations