INDEX

Explanations

characters and numbers from various languages and scripts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 propOrder

-1.21

NameInMap

-0.83

 دیکھیے

-0.82

 Theſe

-0.81

 ainfi

-0.81

paravant

-0.80

StructEnd

-0.78

 AssemblyCulture

-0.75

Personensuche

-0.74

 beginnetje

-0.73

POSITIVE LOGITS

म

0.84

ב

0.83

מ

0.83

ग

0.82

स

0.82

ח

0.82

ब

0.82

פ

0.81

अ

0.81

क

0.81

Activations Density 0.077%

characters and numbers from various languages and scripts

No Comments

No Known Activations