INDEX

Explanations

phrases related to political statements or positions

oai_token-act-pair · gpt-3.5-turbo

periods and end punctuation, indicating sentence conclusions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

utical

-0.73

ĺħ

-0.71

¥µ

-0.64

bernatorial

-0.63

agi

-0.63

uph

-0.62

 Versus

-0.61

aturdays

-0.60

 girlfriends

-0.60

xual

-0.60

POSITIVE LOGITS

Writing

0.93

Scotland

0.92

Britain

0.90

Speaking

0.84

BBC

0.82

Asked

0.82

Labour

0.81

However

0.74

Ireland

0.73

 Ulster

0.69

Activations Density 0.396%

phrases related to political statements or positions

periods and end punctuation, indicating sentence conclusions

No Comments

No Known Activations