INDEX

Explanations

names of famous individuals or public figures

oai_token-act-pair · gpt-3.5-turbo

prominent public figures and their associations or actions

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

Ir

-0.80

Reviewer

-0.79

Ire

-0.79

Maria

-0.71

Irish

-0.69

ãĥĭ

-0.69

Northern

-0.67

FU

-0.66

Native

-0.65

URI

-0.64

POSITIVE LOGITS

steen

0.90

Jr

0.90

agher

0.80

gaard

0.75

Sr

0.75

bey

0.74

III

0.73

 famously

0.72

ovich

0.71

 imperson

0.70

Activations Density 0.152%

names of famous individuals or public figures

prominent public figures and their associations or actions

No Comments

No Known Activations