INDEX

Explanations

the word "hate" at various intensities

oai_token-act-pair · gpt-3.5-turbo

expressions of hatred

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

ItemImage

-0.84

aqu

-0.82

aunder

-0.81

igmatic

-0.80

istics

-0.76

uggest

-0.75

aver

-0.74

DragonMagazine

-0.74

enture

-0.73

arta

-0.73

POSITIVE LOGITS

fully

1.11

 hated

0.93

 hate

0.87

FUL

0.80

 hates

0.79

 wasting

0.79

 Hate

0.78

0.75

hate

0.75

ful

0.75

Activations Density 0.021%

the word "hate" at various intensities

expressions of hatred

No Comments

No Known Activations