INDEX

Explanations

negative or critical words and phrases

oai_token-act-pair · gpt-3.5-turbo

variations of the word "play."

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

âĸ¬

-0.77

HAHA

-0.73

âĸ¬âĸ¬

-0.73

 Logged

-0.68

////////////////////////////////

-0.67

ä¸Ģ

-0.65

fav

-0.63

spin

-0.63

ĪĴ

-0.63

QUI

-0.63

POSITIVE LOGITS

acement

1.25

asma

1.21

atinum

1.17

enty

1.15

umbing

1.05

acements

1.02

icably

1.01

ains

1.00

atoon

0.98

atter

0.98

Activations Density 0.014%

negative or critical words and phrases

variations of the word "play."

No Comments

No Known Activations