INDEX

Explanations

specific mentions of locations within a religious context, particularly highlighting the word "temple."

oai_token-act-pair · gpt-3.5-turbo

references to places of worship, specifically temples

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

BAT

-0.72

Published

-0.72

ÑĮ

-0.71

ï¸

-0.71

-0.67

OVER

-0.67

sen

-0.66

abet

-0.65

 MacArthur

-0.65

 Matter

-0.64

POSITIVE LOGITS

 temple

1.36

 temples

1.22

 altar

0.93

 maiden

0.92

 shrine

0.88

 ceremonies

0.81

mage

0.81

 procession

0.80

 ceremony

0.80

 courtyard

0.79

Activations Density 0.008%

specific mentions of locations within a religious context, particularly highlighting the word "temple."

references to places of worship, specifically temples

No Comments

No Known Activations