INDEX

Explanations

information or content below advertisements in a text

oai_token-act-pair · gpt-3.5-turbo

references to layered content or additional information following a main topic

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

è¦ļéĨĴ

-0.76

venge

-0.65

kowski

-0.64

bags

-0.63

itive

-0.60

 Compan

-0.60

cest

-0.59

Writ

-0.59

 Bows

-0.59

OH

-0.58

POSITIVE LOGITS

 advertisement

0.83

imeo

0.78

 Thumbnails

0.72

 slideshow

0.70

 Pinterest

0.68

 slide

0.62

 Sydney

0.61

 Advertisement

0.61

igon

0.61

atu

0.60

Activations Density 0.011%

information or content below advertisements in a text

references to layered content or additional information following a main topic

No Comments

No Known Activations