INDEX

Explanations

phrases related to proximity or being close in physical or conceptual space

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 Mundi

-0.63

 bodyParser

-0.60

 iParam

-0.59

Bibliographie

-0.57

ismet

-0.54

 ontstaan

-0.53

anjutnya

-0.53

QMetaType

-0.52

ğraf

-0.52

POSITIVE LOGITS

near

0.89

 near

0.86

 Near

0.80

sighted

0.80

 NEAR

0.79

NEAR

0.76

 enough

0.76

Near

0.74

enough

0.66

 resourceCulture

0.63

Activations Density 0.104%

phrases related to proximity or being close in physical or conceptual space

No Comments

No Known Activations