INDEX

Explanations

phrases related to specific movie titles or franchises

oai_token-act-pair · gpt-3.5-turbo

references to franchises or sequels in media titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

rouse

-0.80

estate

-0.74

 privile

-0.72

andan

-0.71

ometimes

-0.70

bably

-0.68

ily

-0.68

cknowled

-0.67

 closely

-0.67

sic

-0.66

POSITIVE LOGITS

 Darkness

1.10

 Empires

1.08

 Machines

0.97

 Atlantis

0.95

 Camel

0.92

 Colossus

0.91

 Wonders

0.91

 Shadows

0.91

 Titans

0.91

 Korra

0.90

Activations Density 0.082%

phrases related to specific movie titles or franchises

references to franchises or sequels in media titles

No Comments

No Known Activations