INDEX

Explanations

mentions of the word "Dom" in a text

oai_token-act-pair · gpt-3.5-turbo

references to the brand 'Domino's'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

WAYS

-0.87

HCR

-0.80

gow

-0.76

phrine

-0.73

packing

-0.73

FUL

-0.72

creen

-0.72

journal

-0.71

GROUND

-0.71

side

-0.70

POSITIVE LOGITS

inant

1.09

ains

1.09

estic

1.02

ino

0.99

ination

0.97

inating

0.90

aine

0.86

ainer

0.86

egu

0.86

aned

0.86

Activations Density 0.007%

mentions of the word "Dom" in a text

references to the brand 'Domino's'

No Comments

No Known Activations