INDEX

Explanations

phrases associated with programming directives and coding terminology

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

Not in Any Lists

No Comments

Negative Logits

لي

-0.66

 нашли

-0.61

vara

-0.61

ettare

-0.58

 брать

-0.58

 Leit

-0.57

бран

-0.57

RTLU

-0.56

no

-0.56

 lega

-0.55

POSITIVE LOGITS

AsUp

1.19

 tranſ

0.88

 raiſ

0.86

 Проци

0.84

brigens

0.83

 خارجية

0.82

 deſt

0.81

 poffe

0.80

 itſelf

0.80

 myſelf

0.77

Activations Density 0.805%

phrases associated with programming directives and coding terminology

No Comments

No Known Activations