INDEX

Explanations

mentions of the word "Jordan"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Jordan."

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

anium

-0.68

 accur

-0.67

 reproduce

-0.66

 Cthulhu

-0.66

 reprodu

-0.66

Â

-0.65

UP

-0.65

Serial

-0.65

 nanop

-0.65

ugen

-0.64

POSITIVE LOGITS

 Jordan

3.87

Jordan

3.13

 Jordanian

2.36

 Petra

1.43

 Jord

1.43

ordan

1.42

 Jamal

1.36

 Lebanon

1.35

 Morocco

1.31

 Jericho

1.29

Activations Density 0.014%

mentions of the word "Jordan"

mentions of the name "Jordan."

No Comments

No Known Activations