INDEX

Explanations

specific references to political terms and figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 39-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_39/width_16k/average_l0_131

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.39.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

 commandement

-0.36

 jugement

-0.36

 Masson

-0.32

 charité

-0.32

 vallée

-0.32

 souverain

-0.31

 clôture

-0.31

 Hoffmann

-0.31

 abus

-0.31

 prati

-0.31

POSITIVE LOGITS

 <=",

0.67

 الحره

0.65

Hentet

0.64

apimachinery

0.60

 &___

0.59

 Chwiliwch

0.59



0.56

Ӕ

0.53

 mít

0.52

intense

0.52

Activations Density 0.133%

specific references to political terms and figures

No Comments

No Known Activations