INDEX

Explanations

references to currency codes

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

 pleaſure

-0.97

 himſelf

-0.91

 houſe

-0.88

 faſt

-0.84

 myſelf

-0.83

 themſelves

-0.83

 purpoſe

-0.82

 ſmall

-0.82

 Monfieur

-0.80

 itſelf

-0.80

POSITIVE LOGITS

 صوتيه

0.85

endpush

0.65

DNEY

0.63

wpi

0.63

AfterClass

0.63

 elegans

0.60

osoever

0.60

getDescriptor

0.59

QUADS

0.58

RenderAtEndOf

0.58

Activations Density 0.008%

references to currency codes

No Comments

No Known Activations