INDEX

Explanations

words related to purchasing and buying actions

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

buy

-0.55

ness

-0.55

cape

-0.55

AndEndTag

-0.54

 Montague

-0.52

 preocupes

-0.52

 watered

-0.51

NESS

-0.51

SCAPE

-0.50

__/

-0.50

POSITIVE LOGITS

 препратки

0.65

 Virol

0.65

 hObject

0.65

tagHelperRunner

0.59

 Geraadpleegd

0.59

endaftar

0.57

fjspx

0.57

sillos

0.57

 الحره

0.56

gemaakt

0.56

Activations Density 0.026%

words related to purchasing and buying actions

No Comments

No Known Activations