INDEX

Explanations

instances of the word "Had" or question words like "Was," "Did," and "Were," suggesting a focus on past actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

was

-0.73

is

-0.72

di

-0.71

in

-0.69

to

-0.69

and

-0.68

de

-0.67

la

-0.64

can

-0.64

con

-0.64

POSITIVE LOGITS

 Monfieur

1.50

Was

1.42

 Diſ

1.32

Been

1.28

 purpoſe

1.28

Efq

1.28

Has

1.27

Would

1.27

ſelf

1.27

 Anſ

1.26

Activations Density 0.143%

instances of the word "Had" or question words like "Was," "Did," and "Were," suggesting a focus on past actions or conditions

No Comments

No Known Activations