INDEX

Explanations

references to work-related tasks or activities

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

+#+#

-0.64

argout

-0.63

SuccessListener

-0.62

RefreshLayout

-0.61

orianCalendar

-0.57

WriteBarrier

-0.55

遺産

-0.55

 fracaso

-0.55

HasBeenSet

-0.54

lå

-0.54

POSITIVE LOGITS

幅

0.59

הל

0.54

声

0.52

örd

0.51

 habari

0.51

onucleotide

0.51

聲

0.51

glio

0.50

fore

0.50

jooq

0.50

Activations Density 0.041%

references to work-related tasks or activities

No Comments

No Known Activations