INDEX

Explanations

the name "Maury" with varying activations

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "ury" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

insula

-0.71

chart

-0.67

eworld

-0.67

hatt

-0.65

 interoper

-0.65

anooga

-0.64

eared

-0.63

 Hilbert

-0.62

verified

-0.61

eway

-0.61

POSITIVE LOGITS

ury

1.36

stein

0.86

uries

0.83

uay

0.75

osity

0.72

jury

0.72

ous

0.69

Å«

0.68

igans

0.67

ry

0.66

Activations Density 0.007%

the name "Maury" with varying activations

mentions of the word "ury" in various contexts

No Comments

No Known Activations