INDEX

Explanations

mentions of newspapers, with a focus on The Washington Post

oai_token-act-pair · gpt-3.5-turbo

references to specific news publications, particularly the Washington Post and the New York Times

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

sed

-0.82

tions

-0.77

 Nadu

-0.76

cles

-0.72

pires

-0.69

pired

-0.68

ardless

-0.67

ĪĴ

-0.66

jab

-0.65

awaru

-0.63

POSITIVE LOGITS

 editorial

1.19

 Magazine

1.19

 Editorial

1.17

 columnist

1.05

 newspaper

1.04

 article

1.00

 bestselling

0.98

 Newspaper

0.97

 Archive

0.94

 Online

0.94

Activations Density 0.081%

mentions of newspapers, with a focus on The Washington Post

references to specific news publications, particularly the Washington Post and the New York Times

No Comments

No Known Activations