INDEX

Explanations

statistics or numerical proportions followed by a category or group of people

oai_token-act-pair · gpt-3.5-turbo

phrases indicating statistical or numerical occurrences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

BO

-0.64

 misunder

-0.61

 clad

-0.61

 streng

-0.57

bye

-0.56

 posing

-0.55

ãĤ´ãĥ³

-0.54

inelli

-0.54

 throats

-0.53

stuff

-0.53

POSITIVE LOGITS

illion

0.85

elfth

0.84

clusively

0.81

pload

0.80

patient

0.79

ousand

0.75

thood

0.74

irteen

0.73

ordinate

0.72

animate

0.68

Activations Density 0.076%

statistics or numerical proportions followed by a category or group of people

phrases indicating statistical or numerical occurrences

No Comments

No Known Activations