INDEX

Explanations

the word "Dick."

oai_token-act-pair · gpt-3.5-turbo

references to the name "Dick" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

yip

-0.76

tics

-0.67

xual

-0.65

ignty

-0.65

 waters

-0.64

ctic

-0.63

bidden

-0.63

hips

-0.62

 Aval

-0.62

 hepatitis

-0.61

POSITIVE LOGITS

enson

1.18

inson

1.13

erson

1.09

 Cheney

1.02

ERSON

0.97

stein

0.96

ie

0.92

minster

0.89

rey

0.88

 Grayson

0.87

Activations Density 0.021%

the word "Dick."

references to the name "Dick" in various contexts

No Comments

No Known Activations