INDEX

Explanations

GIFs in short documents

oai_token-act-pair · gpt-3.5-turbo

references to GIFs

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anwhile

-0.68

lied

-0.68

idem

-0.67

weak

-0.66

ologically

-0.64

kamp

-0.64

arios

-0.63

aos

-0.62

chedel

-0.62

okin

-0.62

POSITIVE LOGITS

GIF

1.37

gif

1.10

TED

0.95

 Thumbnails

0.94

eries

0.83

FIX

0.80

Textures

0.74

LG

0.73

FB

0.71

TF

0.71

Activations Density 0.007%

GIFs in short documents

references to GIFs

No Comments

No Known Activations