INDEX

Explanations

mentions of the word "Swift" with various numerical activations representing different contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of Taylor Swift

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

terior

-0.80

abases

-0.78

berra

-0.73

cheon

-0.73

asher

-0.73

irin

-0.73

itial

-0.72

yrinth

-0.72

awatts

-0.71

ulhu

-0.70

POSITIVE LOGITS

 Swift

0.75

 Swim

0.72

strike

0.72

plain

0.69

song

0.69

 Sparrow

0.68

flake

0.67

KEN

0.67

ipeg

0.66

heart

0.65

Activations Density 0.040%

mentions of the word "Swift" with various numerical activations representing different contexts

mentions of Taylor Swift

No Comments

No Known Activations