INDEX

Explanations

numeric or structured data representations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 يتيمه

-0.63

errHandler

-0.59

 Numerade

-0.56

khe

-0.56

+#+#

-0.56

 متعلقه

-0.56

readObject

-0.54

 للمعارف

-0.54

createObject

-0.53

Zeneca

-0.52

POSITIVE LOGITS

Personendaten

0.61

webElement

0.60

Enllaces

0.51

landen

0.50

PerformLayout

0.50

 umfass

0.50

 voudrais

0.48

 Roskov

0.47

 Comparator

0.47

Cyfeiriadau

0.46

Activations Density 0.070%

numeric or structured data representations

No Known Activations