INDEX

Explanations

references to locations and transformations within spaces

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

Iné

-0.30

 właścic

-0.29

 dueños

-0.29

 Geset

-0.29

empereur

-0.28

 čierna

-0.28

 arché

-0.28

Enlaces

-0.28

 propriétaires

-0.28

 Fuchs

-0.27

POSITIVE LOGITS

 Савезне

0.71

Попис

0.65

IVEREF

0.60

 beſte

0.60

URLException

0.57

converting

0.56

 conversions

0.56

 Utili

0.56

 converting

0.56

jgl

0.55

Activations Density 0.309%

references to locations and transformations within spaces

No Comments

No Known Activations