INDEX

Explanations

mentions of former positions or titles

oai_token-act-pair · gpt-3.5-turbo

the term "Former" followed by a title or role

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

otle

-0.94

anguage

-0.84

achus

-0.82

antics

-0.78

acht

-0.72

andise

-0.72

okin

-0.72

matic

-0.72

aceae

-0.71

thora

-0.70

POSITIVE LOGITS

 Yugoslav

0.99

 Yugoslavia

0.96

 President

0.84

 Soviet

0.83

 presidents

0.83

 Deputy

0.78

 president

0.78

 colleague

0.78

bies

0.76

 captives

0.76

Activations Density 0.035%

mentions of former positions or titles

the term "Former" followed by a title or role

No Comments

No Known Activations