INDEX

Explanations

phrases related to legal proceedings and potential misconduct

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

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

ftagPool

-0.52

colhead

-0.49

 Vikipedi

-0.49

﴿

-0.47

لار

-0.47

 Audiodateien

-0.46

InputModule

-0.46

istocene

-0.46

سلحة

-0.45

 entendre

-0.45

POSITIVE LOGITS

 autorytatywna

0.75

 gepubliceerd

0.66

 <<<<<<<<<<<<<<

0.65

NameInMap

0.65

 ISNI

0.63

Controllo

0.56

Autoritní

0.55

новништво

0.54

tiéndose

0.54

 Taktlose

0.54

Activations Density 1.088%

phrases related to legal proceedings and potential misconduct

No Comments

No Known Activations