INDEX

Explanations

references to public opinion and voting behavior

oai_token-act-pair · gpt-3.5-turbo

phrases and words related to public opinion and societal issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 aliases

-0.68

ortment

-0.66

 promotion

-0.65

lic

-0.63

 delet

-0.63

 Xuan

-0.63

 Equip

-0.62

ourney

-0.62

tein

-0.62

iba

-0.61

POSITIVE LOGITS

 overwhelmingly

1.20

 clam

1.12

 vote

1.05

 distrust

1.01

 revolt

0.98

 perceive

0.94

 Vote

0.93

 mistrust

0.93

 electing

0.92

 forgive

0.89

Activations Density 0.458%

references to public opinion and voting behavior

phrases and words related to public opinion and societal issues

No Comments

No Known Activations