INDEX

Explanations

references to the completion or production of items or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

 touristes

-0.60

 religieuses

-0.58

 Tourists

-0.57

 confirmer

-0.56

 mierda

-0.56

 vectorielles

-0.56

qtype

-0.55

approximately

-0.55

whereas

-0.54

 electors

-0.54

POSITIVE LOGITS

hysema

0.63

AccessFile

0.61

DockStyle

0.59

CloseOperation

0.59

''

0.57

ngOn

0.56

 gepubliceerd

0.54

kwds

0.54

ждународ

0.53

ochlor

0.53

Activations Density 0.023%

references to the completion or production of items or processes

No Comments

No Known Activations