INDEX

Explanations

Proper nouns, particularly names of people and organizations

oai_token-act-pair · gpt-3.5-turbo

sections that describe different Pokémon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

è¦ļéĨĴ

-0.72

 Metatron

-0.72

 Ribbon

-0.67

 thereof

-0.66

lit

-0.64

 mileage

-0.63

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.63

åŃĲ

-0.63

 guiName

-0.62

 sizeof

-0.61

POSITIVE LOGITS

jamin

1.21

odore

1.19

alyst

1.17

lahoma

1.15

hesda

1.13

ropolitan

1.11

resa

1.10

ifax

1.01

dinand

1.00

mond

1.00

Activations Density 0.260%

Proper nouns, particularly names of people and organizations

sections that describe different Pokémon

No Comments

No Known Activations