INDEX

Explanations

countries, organizations, and specific actions or situations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating restrictions or limitations

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

 Seym

-0.76

 Vaugh

-0.67

 thous

-0.62

ãĤ´ãĥ³

-0.61

 mathemat

-0.56

 nodd

-0.54

 Kardash

-0.53

Math

-0.53

anwhile

-0.52

edIn

-0.52

POSITIVE LOGITS

geon

0.65

hee

0.60

ree

0.59

brew

0.54

Ĥİ

0.53

Ī

0.52

Ĥª

0.51

TOR

0.50

ndra

0.50

aves

0.49

Activations Density 0.611%

countries, organizations, and specific actions or situations

phrases indicating restrictions or limitations

No Comments

No Known Activations