INDEX

Explanations

occurrences of fullness or content-related descriptors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

的下

-0.41

ług

-0.41

ourge

-0.41

 press

-0.41

Sea

-0.40

ைக்

-0.40

piram

-0.40

átka

-0.39

!("{}",

-0.39

 Without

-0.39

POSITIVE LOGITS

SequentialGroup

0.81

eload

0.81

 contenir

0.79

ंदीखरीदारी

0.78

expandindo

0.75

Personendaten

0.74

writerow

0.72

 '\\;'

0.71

 berisi

0.71

 rempli

0.70

Activations Density 0.346%

occurrences of fullness or content-related descriptors

No Comments

No Known Activations