INDEX

Explanations

specific mentions of the term "Indy" alongside other related words or phrases

oai_token-act-pair · gpt-3.5-turbo

references to the "IndyCar" racing series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Gleaming

-0.84

aunder

-0.79

ascus

-0.73

-0.72

istics

-0.72

utenberg

-0.71

ILA

-0.70

owers

-0.69

espie

-0.69

atche

-0.69

POSITIVE LOGITS

 Indy

1.03

 Eleven

1.01

Car

0.77

 Lights

0.77

 Pacers

0.75

 Apost

0.73

apolis

0.71

 Races

0.70

 Pole

0.69

 Libre

0.68

Activations Density 0.008%

specific mentions of the term "Indy" alongside other related words or phrases

references to the "IndyCar" racing series

No Comments

No Known Activations