INDEX

Explanations

mentions of locations, specifically ones starting with "Fort"

oai_token-act-pair · gpt-3.5-turbo

references to specific military forts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Magikarp

-0.80

gger

-0.74

lihood

-0.71

lyak

-0.67

FREE

-0.64

Companies

-0.63

stakes

-0.63

Publisher

-0.62

OPLE

-0.61

 subscribing

-0.61

POSITIVE LOGITS

uitous

1.37

itude

1.17

 Fort

1.09

ieth

1.03

 Lauderdale

1.00

oise

0.94

dinand

0.93

uit

0.91

folios

0.90

ilage

0.89

Activations Density 0.007%

mentions of locations, specifically ones starting with "Fort"

references to specific military forts

No Comments

No Known Activations