INDEX

Explanations

words and phrases indicating importance or intensity

oai_token-act-pair · gemini-2.0-flash

high concentrations

np_token-act-pair-logits · gpt-4o-mini

high

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 BoxDecoration

-0.93

RectangleBorder

-0.88

AnchorStyles

-0.83

ChildScrollView

-0.79

httphttps

-0.79

 defaultstate

-0.76

Referencies

-0.75

UnsafeEnabled

-0.75

Personensuche

-0.74

 ModelExpression

-0.74

POSITIVE LOGITS

 concentration

0.56

 concentrations

0.55

but

0.49

 resolve

0.47

 percentage

0.44

yat

0.43

 concentrate

0.43

 Hotel

0.43

 quality

0.43

 concent

0.43

Activations Density 4.366%

words and phrases indicating importance or intensity

high concentrations

high

No Comments

No Known Activations