INDEX

Explanations

references to a specific news outlet "The Washington Post."

oai_token-act-pair · gpt-3.5-turbo

mentions of "The Washington Post."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

hra

-0.83

vana

-0.83

avorite

-0.78

netflix

-0.73

acca

-0.72

thood

-0.68

merce

-0.67

iors

-0.66

SELECT

-0.66

addafi

-0.66

POSITIVE LOGITS

 Times

1.07

 Herald

1.04

 Examiner

1.04

 Observer

0.97

 Post

0.91

 Journal

0.91

 Evening

0.90

 Blade

0.90

 Massacre

0.88

 Morning

0.87

Activations Density 0.082%

references to a specific news outlet "The Washington Post."

mentions of "The Washington Post."

No Comments

No Known Activations