INDEX

Explanations

mentions of the name "Theresa."

oai_token-act-pair · gpt-3.5-turbo

references to Theresa May

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

assian

-0.93

ilver

-0.78

erve

-0.70

oned

-0.69

 Helsinki

-0.68

 Kard

-0.68

displayText

-0.67

oning

-0.67

stood

-0.67

dash

-0.66

POSITIVE LOGITS

 Theresa

1.17

 Thatcher

0.77

resa

0.74

 Andrea

0.64

 Rebell

0.63

 Wynne

0.62

nuts

0.62

borough

0.62

 divorce

0.61

lehem

0.61

Activations Density 0.006%

mentions of the name "Theresa."

references to Theresa May

No Comments

No Known Activations