INDEX

Explanations

proper nouns related to political figures

oai_token-act-pair · gpt-3.5-turbo

references to various Prime Ministers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

anwhile

-0.91

bleacher

-0.76

entric

-0.76

okia

-0.75

ologies

-0.72

avery

-0.71

pel

-0.71

schild

-0.71

 dotted

-0.70

oso

-0.68

POSITIVE LOGITS

 Minister

1.14

 ministerial

1.01

 Ministers

0.95

 minister

0.91

 ministers

0.86

 Office

0.82

 Election

0.81

Rib

0.80

 Affairs

0.78

 Prime

0.77

Activations Density 0.012%

proper nouns related to political figures

references to various Prime Ministers

No Comments

No Known Activations