INDEX

Explanations

the word "Princess"

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Princess" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 spaced

-0.67

 smoker

-0.66

NRS

-0.66

OUT

-0.66

sych

-0.66

umbn

-0.65

ORN

-0.64

EFF

-0.63

ucl

-0.63

 revolving

-0.63

POSITIVE LOGITS

 Leia

1.13

 Princess

1.11

 Bride

1.07

 Peach

0.95

 Diana

0.91

anova

0.90

 Celest

0.84

 princess

0.83

 Toad

0.82

 Royale

0.79

Activations Density 0.008%

the word "Princess"

mentions of the word "Princess" in various contexts

No Comments

No Known Activations