INDEX

Explanations

terms related to virtual entities or qualities

oai_token-act-pair · gpt-3.5-turbo

the word "Virt" and its variations, suggesting a focus on the concept of virtue or related topics

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

dress

-0.74

PLIED

-0.69

Interstitial

-0.66

--------------------------------------------------------

-0.66

corn

-0.65

printed

-0.65

 oats

-0.62

ONT

-0.62

seller

-0.61

 dividing

-0.61

POSITIVE LOGITS

 Virt

1.28

uous

1.01

ually

0.94

virt

0.92

atility

0.85

inguished

0.84

ues

0.83

 virt

0.83

cipled

0.81

Scroll

0.80

Activations Density 0.011%

terms related to virtual entities or qualities

the word "Virt" and its variations, suggesting a focus on the concept of virtue or related topics

No Comments

No Known Activations