INDEX

Explanations

proper nouns associated with people, specifically in a variety of contexts

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

names of people

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

ViewInit

-0.55

AddTagHelper

-0.53

%)$

-0.53

$?

-0.52

AccessorTable

-0.51

"])

-0.50

%")

-0.49

')));

-0.49

]]);

-0.49

//}

-0.49

POSITIVE LOGITS

 zijne

0.63

 redenen

0.53

 temprana

0.53

 McKee

0.52

 plegable

0.51

 vectorielles

0.50

 nødven

0.50

 betrekking

0.49

 vectorielle

0.49

StackNavigator

0.49

Activations Density 0.375%

proper nouns associated with people, specifically in a variety of contexts

names of people

No Comments

No Known Activations