INDEX

Explanations

phrases indicating strong emotions or confrontations

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "fired back at" or similar confrontational phrases

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

Cert

-0.75

glass

-0.74

BuyableInstoreAndOnline

-0.74

FORMATION

-0.73

Lens

-0.72

mingham

-0.70

photo

-0.70

fill

-0.70

phone

-0.66

widget

-0.65

POSITIVE LOGITS

 least

1.23

 halftime

0.88

rial

0.86

onement

0.81

 midnight

0.80

rium

0.80

 dusk

0.79

 midday

0.76

roph

0.76

ĸļ

0.75

Activations Density 0.179%

phrases indicating strong emotions or confrontations

instances of the phrase "fired back at" or similar confrontational phrases

No Comments

No Known Activations