INDEX

Explanations

mentions of males or men

oai_token-act-pair · gpt-3.5-turbo

references to men

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

inventoryQuantity

-0.69

burst

-0.67

Layer

-0.64

ID

-0.64

ellation

-0.64

kick

-0.63

estone

-0.63

 Collider

-0.62

arthed

-0.61

balance

-0.61

POSITIVE LOGITS

men

3.52

Men

2.41

Men

2.39

 males

2.23

 women

2.14

 boys

2.09

men

1.95

MEN

1.92

 dudes

1.85

 gentlemen

1.81

Activations Density 0.031%

mentions of males or men

references to men

No Comments

No Known Activations