INDEX

Explanations

phrases that express knowledge or understanding, particularly in relation to personal assessments and opinions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

as far as I know

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

Negative Logits

 istället

-0.29

 привод

-0.28

$_['

-0.28

 ekst

-0.28

 fallait

-0.27

ergic

-0.27

 едино

-0.27

leştir

-0.27

 dési

-0.27

 often

-0.26

POSITIVE LOGITS

 BorderSide

0.66

jmniej

0.65

</tfoot>

0.60

StoreMessageInfo

0.60

GEBURTSDATUM

0.60

 trouw

0.59

OGND

0.59

Cartney

0.57

 Ahnung

0.56

kegaard

0.55

Activations Density 0.008%

phrases that express knowledge or understanding, particularly in relation to personal assessments and opinions

as far as I know

No Known Activations