INDEX

Explanations

structured formats for questions and lists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 6-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_6/width_16k/average_l0_133

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.6.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

armée

-0.45

Carriera

-0.43

rayage

-0.43

 serie

-0.42

hene

-0.42

 cruda

-0.42

='',

-0.41

хьтан

-0.41

cèse

-0.40

efel

-0.40

POSITIVE LOGITS

SBATCH

0.90

InputTagHelper

0.88

 autorytatywna

0.82

abestanden

0.72

 postIndex

0.71

 CreateTagHelper

0.68

存于互联网档案馆

0.66

GEBURTSDATUM

0.62

ResumeLayout

0.59

 becauſe

0.59

Activations Density 0.205%

structured formats for questions and lists

No Comments

No Known Activations