INDEX

Explanations

specific names and proper nouns

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_131k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.33.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

-0.60

-0.53

колай

-0.44

ی

-0.44

ا

-0.44

jago

-0.44

lyl

-0.43

er

-0.43

uzz

-0.43

olia

-0.42

POSITIVE LOGITS

 resourceCulture

0.69

Personendaten

0.66

Życiorys

0.54

 propOrder

0.54

 насељу

0.53

DockStyle

0.52

ity

0.50

XOXO

0.49

EndInit

0.49

jší

0.48

Activations Density 0.035%

specific names and proper nouns

No Comments

No Known Activations