INDEX

Explanations

commands and expressions of frustration or authority

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

دانشنامهٔ

-0.47

Попис

-0.47

Искәрмәләр

-0.42

TargetApi

-0.42

BeforeMethod

-0.41

Życiorys

-0.40

はいけない

-0.39

ChildScrollView

-0.39

évaluateur

-0.38

といけない

-0.38

POSITIVE LOGITS

Shut

0.81

 SHUT

0.80

 stop

0.79

 shut

0.79

shut

0.76

 Shut

0.75

Stop

0.71

 Stop

0.70

🖕

0.69

 cease

0.68

Activations Density 0.327%

commands and expressions of frustration or authority

No Comments

No Known Activations