INDEX

Explanations

mentions of the name "Catherine"

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly "Catherine" and "Sebastian"

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

 wrapper

-0.71

NX

-0.67

 kernels

-0.66

WD

-0.66

:\

-0.65

 nuts

-0.64

 Loop

-0.63

 Terminal

-0.63

HI

-0.62

 True

-0.62

POSITIVE LOGITS

 Catherine

2.19

Peter

1.93

 POLITICO

1.56

atherine

1.52

 Piet

1.39

ilda

1.38

 Peter

1.34

 Sebastian

1.33

 Wang

1.31

 Katherine

1.22

Activations Density 0.012%

mentions of the name "Catherine"

mentions of specific names, particularly "Catherine" and "Sebastian"

No Comments

No Known Activations