INDEX

Explanations

titles or roles related to different professions or positions

oai_token-act-pair · gpt-3.5-turbo

specific job titles or roles

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

ilion

-0.91

etting

-0.80

heet

-0.72

inki

-0.70

esting

-0.68

urgical

-0.68

perm

-0.68

oho

-0.67

liga

-0.67

creen

-0.66

POSITIVE LOGITS

 extraord

1.29

hood

1.10

esses

0.97

 Beware

0.91

 beware

0.87

stown

0.85

NAME

0.85

Joined

0.84

ess

0.83

ially

0.79

Activations Density 0.406%

titles or roles related to different professions or positions

specific job titles or roles

No Comments

No Known Activations