INDEX

Explanations

terms of address or respectful language directed towards authority figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 himſelf

-0.66

 Monfieur

-0.64

DockStyle

-0.59

дент

-0.58

 endforeach

-0.57

 CreateTagHelper

-0.57

romptu

-0.56

 Spaniards

-0.56

 переписи

-0.56

 poffible

-0.56

POSITIVE LOGITS

 excell

0.79

 Highness

0.74

HIS

0.71

 Majesty

0.70

 Excellency

0.69

 Honor

0.69

HIS

0.68

His

0.68

 Excell

0.68

His

0.67

Activations Density 0.098%

terms of address or respectful language directed towards authority figures

No Comments

No Known Activations