INDEX

Explanations

articles or content related to news events or stories

oai_token-act-pair · gpt-3.5-turbo

instances of the word "RELATED" and other similar labels or tags in the text

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

stood

-0.85

-0.78

angers

-0.73

amping

-0.71

apers

-0.71

udi

-0.70

atur

-0.70

animate

-0.70

ctrl

-0.68

oise

-0.68

POSITIVE LOGITS

 VIDEOS

1.13

 IMAGES

1.13

 INFORMATION

1.08

 STOR

0.97

RELATED

0.96

 ARTICLE

0.95

 STORY

0.92

 LINK

0.88

...]

0.87

 ALSO

0.87

Activations Density 0.011%

articles or content related to news events or stories

instances of the word "RELATED" and other similar labels or tags in the text

No Comments

No Known Activations