INDEX

Explanations

film-related entities such as movie titles, actors, and production information

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, places, and proper nouns

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tooth

-0.71

 accelerated

-0.71

 forgiven

-0.69

 reintrodu

-0.66

 worrying

-0.65

ãĢİ

-0.64

 tightening

-0.64

 perfect

-0.64

 implementing

-0.64

 setting

-0.63

POSITIVE LOGITS

(@

0.95

cerpt

0.87

GoldMagikarp

0.86

anan

0.80

maxwell

0.77

Wolf

0.77

idav

0.77

ascript

0.76

Anonymous

0.76

veyard

0.75

Activations Density 0.281%

film-related entities such as movie titles, actors, and production information

mentions of specific names, places, and proper nouns

No Comments

No Known Activations