INDEX

Explanations

references to programming or computer-related terminology

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

OrWhiteSpace

-0.66

 bezeichneter

-0.65

 program

-0.61

})();

-0.60

 oprot

-0.60

()]

-0.59

newArrayList

-0.59

/>";

-0.57

XDECREF

-0.56

 للمعارف

-0.55

POSITIVE LOGITS

الإنجليزية

0.59

roxene

0.56

 quæ

0.56

HomeAsUpEnabled

0.54

 Bucure

0.53

Autoritní

0.52

 diarrhoea

0.52

melhor

0.52

NamedQueries

0.51

CodedInputStream

0.50

Activations Density 0.037%

references to programming or computer-related terminology

No Comments

No Known Activations