INDEX

Explanations

high-level technical terms related to image resolution and pixel manipulation

oai_token-act-pair · gpt-3.5-turbo

terms related to pixels and pixel density in digital images

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

shore

-0.74

avanaugh

-0.74

doms

-0.68

rament

-0.67

enture

-0.67

kson

-0.66

OME

-0.66

redo

-0.65

IFE

-0.65

UNE

-0.64

POSITIVE LOGITS

ixels

0.88

 pixel

0.87

 shader

0.86

 pixels

0.85

 shading

0.83

ated

0.81

agraph

0.76

 sprites

0.72

acet

0.71

 IMAGES

0.70

Activations Density 0.010%

high-level technical terms related to image resolution and pixel manipulation

terms related to pixels and pixel density in digital images

No Comments

No Known Activations