INDEX

Explanations

phrases related to different professions or demographics describing specific attributes or actions

oai_token-act-pair · gpt-3.5-turbo

references to different professions or roles in society

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 divisions

-0.80

ories

-0.78

 Monsters

-0.75

 Horses

-0.73

bows

-0.72

itars

-0.71

acements

-0.71

 Reps

-0.71

 giants

-0.69

 Dragons

-0.69

POSITIVE LOGITS

who

1.00

 knows

0.92

who

0.87

 learns

0.85

 chooses

0.83

 beware

0.81

 understands

0.80

 instinctively

0.80

hood

0.78

 lear

0.75

Activations Density 0.235%

phrases related to different professions or demographics describing specific attributes or actions

references to different professions or roles in society

No Comments

No Known Activations