INDEX

Explanations

technical terms and procedures related to data and analysis

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

 Nowadays

-0.61

Nowadays

-0.58

 nowadays

-0.57

Moreover

-0.52

 Moreover

-0.51

 shrimps

-0.51

ChildScrollView

-0.50

 softwares

-0.48

 Hongkong

-0.48

 respectivas

-0.45

POSITIVE LOGITS

Datuak

0.73

checkFor

0.65

 preventative

0.63

 ddelwed

0.60

 CreateTagHelper

0.60

arned

0.56

brot

0.54

 Photocase

0.53

~*~

0.53

 يتيمه

0.53

Activations Density 4.142%

technical terms and procedures related to data and analysis

No Comments

No Known Activations