INDEX

Explanations

references to visual imagery and descriptive details in a scene

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

 Comunicação

-0.42

 mencer

-0.38

 renseignements

-0.35

 riego

-0.34

 okuyayım

-0.32

 humedad

-0.32

 idéia

-0.32

 talleres

-0.32

wußt

-0.31

 renouvel

-0.31

POSITIVE LOGITS

 ModelExpression

0.73

 clips

0.65

 tartan

0.62

 propOrder

0.58

 grinning

0.58

 turban

0.58

 selfies

0.57

 montage

0.57

 selfie

0.57

 machete

0.57

Activations Density 0.189%

references to visual imagery and descriptive details in a scene

No Comments

No Known Activations