INDEX

Explanations

phrases related to high levels of excitement or interest surrounding a person, event, or topic

oai_token-act-pair · gpt-3.5-turbo

mentions of hype and its associated contexts

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

tein

-0.70

itaire

-0.66

angan

-0.63

 Graveyard

-0.62

 Sheen

-0.62

 illegally

-0.62

 Buffer

-0.59

 crossings

-0.59

itars

-0.58

Ide

-0.57

POSITIVE LOGITS

 hype

1.23

 tremend

0.91

leading

0.82

hyp

0.79

aic

0.79

 frenzy

0.75

sters

0.74

bub

0.74

rites

0.74

rous

0.73

Activations Density 0.016%

phrases related to high levels of excitement or interest surrounding a person, event, or topic

mentions of hype and its associated contexts

No Comments

No Known Activations