INDEX

Explanations

references to academic journal articles or papers

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

 Farr

-0.39

 Sharpe

-0.37

anda

-0.37

 Leinwand

-0.34

 incessantly

-0.33

狽

-0.33

 impact

-0.32

 meticulously

-0.31

 Gole

-0.31

rend

-0.31

POSITIVE LOGITS

Personendaten

0.78

 autorytatywna

0.76

OGND

0.68

 ModelExpression

0.65

曖昧さ回避

0.63

webElementXpaths

0.62

хьтан

0.59

Personensuche

0.59

NameInMap

0.59

ьаж

0.58

Activations Density 0.115%

references to academic journal articles or papers

No Comments

No Known Activations