INDEX

Explanations

out informations or uncovering details

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the process of discovery or revelation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

oulder

-0.71

aving

-0.69

cius

-0.69

iets

-0.68

asus

-0.67

cious

-0.66

idity

-0.65

orously

-0.63

shaw

-0.63

Textures

-0.61

POSITIVE LOGITS

posts

0.84

 è£ıè

0.83

casts

0.77

skirts

0.73

fitted

0.71

lier

0.71

doors

0.70

stadt

0.69

tical

0.68

wards

0.66

Activations Density 0.051%

out informations or uncovering details

phrases indicating the process of discovery or revelation

No Comments

No Known Activations