INDEX

Explanations

unexpected or surprising events or developments

oai_token-act-pair · gpt-3.5-turbo

phrases indicating unexpected changes or developments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ufact

-0.71

league

-0.69

RES

-0.68

leased

-0.66

particip

-0.65

 Ducks

-0.65

audi

-0.64

hammad

-0.63

 Lama

-0.62

alty

-0.62

POSITIVE LOGITS

 twist

1.24

 twists

1.18

 Twist

0.89

 endings

0.85

 twisting

0.83

 Whedon

0.78

 weave

0.76

stroke

0.73

abouts

0.71

creen

0.71

Activations Density 0.013%

unexpected or surprising events or developments

phrases indicating unexpected changes or developments

No Comments

No Known Activations