INDEX

Explanations

mentions of the name "Yoda."

oai_token-act-pair · gpt-3.5-turbo

references to the character Yoda from the Star Wars franchise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 peer

-0.66

fully

-0.64

 supp

-0.63

 frequency

-0.63

 construct

-0.62

 Merc

-0.62

 Loving

-0.61

 control

-0.61

 Speed

-0.60

 forward

-0.60

POSITIVE LOGITS

oda

4.83

ods

1.31

od

1.27

odan

1.24

odo

1.12

odium

1.11

odi

1.09

uda

1.09

ota

1.09

aga

1.09

Activations Density 0.015%

mentions of the name "Yoda."

references to the character Yoda from the Star Wars franchise

No Comments

No Known Activations