INDEX

Explanations

various forms and configurations of objects or concepts

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

form of

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

<?

-0.52

 فريبيس

-0.51

 PeEnEo

-0.48

+#+

-0.46

ezy

-0.45

RenderAtEndOf

-0.44

enchymal

-0.42

 resourceCulture

-0.40

 Италијани

-0.40

 dosta

-0.39

POSITIVE LOGITS

forms

0.59

form

0.59

 Form

0.57

alised

0.52

Form

0.50

 LoginForm

0.50

 FORM

0.50

 form

0.49

factor

0.49

 Forms

0.49

Activations Density 0.037%

various forms and configurations of objects or concepts

form of

No Comments

No Known Activations