INDEX

Explanations

instances of the word "threaten" or its variations in various contexts indicating risk or danger

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

LabelTagHelper

-0.58

 purpoſe

-0.54

 deſt

-0.52

 juſ

-0.52

 straighten

-0.51

 Anſ

-0.50

 AttributeSet

-0.48

 pinulongan

-0.47

 étoient

-0.47

 Conform

-0.47

POSITIVE LOGITS

div

0.60

 intptr

0.55

 threatened

0.53

div

0.51

Grid

0.50

 threaten

0.50

))/(

0.48

 Democratic

0.47

{?}

0.47

 threatens

0.46

Activations Density 0.204%

instances of the word "threaten" or its variations in various contexts indicating risk or danger

No Comments

No Known Activations