INDEX

Explanations

terms related to disabling or reducing functionality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_15/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.15.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

 AssemblyTitle

-0.65

 QtCore

-0.63

AnchorStyles

-0.62

PageFactory

-0.59

########.

-0.58

ViewInit

-0.56

ollectionView

-0.56

EndInit

-0.56

 bulat

-0.55

 BoxFit

-0.55

POSITIVE LOGITS

Bad

0.71

BAD

0.68

bad

0.67

 disable

0.65

Bad

0.65

disable

0.61

BAD

0.60

bad

0.59

 disabled

0.59

 Disable

0.59

Activations Density 0.600%

terms related to disabling or reducing functionality

No Comments

No Known Activations