INDEX

Explanations

references to people, places, and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ves

-0.58

els

-0.52

SharedCtor

-0.51

ще

-0.49

iment

-0.48

erais

-0.46

чер

-0.46

inputType

-0.46

')}}"

-0.45

 range

-0.45

POSITIVE LOGITS

出版年

0.83

GraphicsUnit

0.80

 виправивши

0.78

 autorytatywna

0.74

 الحره

0.73

findpost

0.73

httphttps

0.73

السكان

0.69

rungsseite

0.68

رشف

0.68

Activations Density 1.006%

references to people, places, and organizations

No Comments

No Known Activations