INDEX

Explanations

titles and proper nouns related to specific events or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 queſta

-0.75

httphttps

-0.74

afficheront

-0.72

 informée

-0.71

webElementXpaths

-0.71

tagHelperRunner

-0.70

 surla

-0.69

ロウィン

-0.66

utilisons

-0.65

awtextra

-0.65

POSITIVE LOGITS

 also

0.34

!="")

0.31

 live

0.30

 overall

0.30

 That

0.30

domain

0.28

 же

0.28

 Jehová

0.28

 then

0.28

 satisfy

0.27

Activations Density 1.489%

titles and proper nouns related to specific events or categories

No Comments

No Known Activations