INDEX

Explanations

sequences of repeated underscores or patterns indicating structured formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 oprot

-0.57

clonal

-0.54

calyptic

-0.53

 Grüsse

-0.52

 Auss

-0.50

ondissement

-0.48

 Cabral

-0.48

прочем

-0.46

 camiseta

-0.45

 cytokines

-0.45

POSITIVE LOGITS

Hozzáférés

1.03

 InputDecoration

0.83

rungsseite

0.79

 autorytatywna

0.73

 насељу

0.71

autogui

0.68

InputBorder

0.67

__':

0.66

 vVar

0.65

Diwedd

0.64

Activations Density 0.051%

sequences of repeated underscores or patterns indicating structured formatting

No Comments

No Known Activations