INDEX

Explanations

references to individuals with previous titles or positions of authority

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Marburg

-0.76

 épis

-0.72

 sneeze

-0.69

 Trish

-0.65

 Kull

-0.64

стоин

-0.64

 fubject

-0.64

 Sien

-0.63

 kwamba

-0.61

IsContent

-0.60

POSITIVE LOGITS

 former

1.24

 Former

1.23

Former

1.19

former

1.01

 ehemalige

0.98

 ehemaligen

0.97

ex

0.95

Formerly

0.94

 Formerly

0.90

ancien

0.89

Activations Density 0.070%

references to individuals with previous titles or positions of authority

No Comments

No Known Activations