INDEX

Explanations

references to collective experiences or actions performed by groups, particularly using "we"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 BoxDecoration

-0.48

 Disease

-0.46

waitUntil

-0.45

Datuak

-0.45

辕

-0.45

antMatchers

-0.43

toMatchSnapshot

-0.43

 Мексичка

-0.43

 CGRectMake

-0.42

المناصب

-0.41

POSITIVE LOGITS

 often

0.50

MessageOf

0.49

<?

0.48

 oftentimes

0.48

 amarillas

0.47

 also

0.47

MessageTagHelper

0.46

 nahilalakip

0.45

 considérons

0.44

 París

0.43

Activations Density 0.602%

references to collective experiences or actions performed by groups, particularly using "we"

No Comments

No Known Activations