INDEX

Explanations

terms related to unique identifiers or repeated instances in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

+#+

-0.70

يكب

-0.68

makeConstraints

-0.63

 للاسماء

-0.60

setVerticalGroup

-0.60

 InputDecoration

-0.60

esez

-0.59

########.

-0.56

PRNewswire

-0.55

<bos>

-0.54

POSITIVE LOGITS

NamedQueries

0.57

 odkazy

0.57

 thread

0.53

ródło

0.52

 threads

0.51

lances

0.50

Хро

0.50

 bones

0.50

 foundation

0.47

 bienvenue

0.47

Activations Density 0.335%

terms related to unique identifiers or repeated instances in context

No Comments

No Known Activations