INDEX

Explanations

terms that indicate relationships or representations within scientific and technical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 works

-0.54

ton

-0.53

Works

-0.51

Uses

-0.50

 interven

-0.48

use

-0.47

use

-0.46

 Works

-0.45

 Caron

-0.45

foot

-0.43

POSITIVE LOGITS

 illustrate

1.47

 illustrates

1.40

 indicate

1.37

 indicates

1.33

indicates

1.26

 represents

1.24

 Indicates

1.24

 imply

1.23

indicate

1.20

 Indicate

1.20

Activations Density 0.233%

terms that indicate relationships or representations within scientific and technical contexts

No Comments

No Known Activations