INDEX

Explanations

phrases that signify indication or suggestion

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

 indicate

-1.52

indicate

-1.14

 indicates

-1.05

 indicating

-0.98

 indicar

-0.87

 indicando

-0.84

 Indicate

-0.83

 indicated

-0.82

 indican

-0.78

 indication

-0.76

POSITIVE LOGITS

Portale

0.75

 mids

0.69

aragua

0.66

 Meld

0.66

 Connectez

0.64

BeginContext

0.64

 eletr

0.63

leuchte

0.63

 WebDriverWait

0.63

encre

0.63

Activations Density 0.037%

phrases that signify indication or suggestion

No Comments

No Known Activations