INDEX

Explanations

terms and phrases related to programming and algorithms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

seguida

-0.54

 nenhuma

-0.53

 nadie

-0.51

 nenhum

-0.51

 ninguno

-0.50

เดียว

-0.49

 dispuesto

-0.48

раздо

-0.48

PostExecute

-0.48

 aucune

-0.47

POSITIVE LOGITS

 voluntarily

0.56

 <<<<<<<<<<<<<<

0.51

mos

0.50

 silently

0.47

apiClient

0.46

 getEmail

0.44

 secretly

0.41

 itſelf

0.41

 getName

0.41

@[+][

0.41

Activations Density 0.119%

terms and phrases related to programming and algorithms

No Comments

No Known Activations