INDEX

Explanations

proper nouns or names, specifically the name "Stone."

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "Stone."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 monop

-0.76

 influenza

-0.75

orial

-0.75

<+

-0.74

 actu

-0.73

owners

-0.72

 caps

-0.70

 comm

-0.70

 commute

-0.68

ANG

-0.66

POSITIVE LOGITS

 Stone

3.91

Stone

2.81

 Stones

2.42

 stone

2.07

stone

1.67

 stones

1.67

stones

1.50

estone

1.28

 Marble

1.28

 Bone

1.22

Activations Density 0.010%

proper nouns or names, specifically the name "Stone."

repeated mentions of the word "Stone."

No Comments

No Known Activations