INDEX

Explanations

articles followed by words indicating a position in a hierarchy or in a physical space

oai_token-act-pair · gemini-2.0-flash

ordinal numbers

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_24/width_16k/average_l0_37

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 most

-0.99

Most

-0.88

most

-0.83

 Most

-0.79

 MOST

-0.73

 EnglishChoose

-0.73

 moft

-0.71

ezers

-0.69

UnsafeEnabled

-0.67

 meest

-0.65

POSITIVE LOGITS

 second

1.17

second

1.06

third

0.98

 fifth

0.97

 third

0.97

 fourth

0.96

sixth

0.94

 sixth

0.93

fifth

0.90

fourth

0.90

Activations Density 2.812%

articles followed by words indicating a position in a hierarchy or in a physical space

ordinal numbers

No Comments

No Known Activations