INDEX

Explanations

places or proper nouns related to specific countries and locations

oai_token-act-pair · gpt-3.5-turbo

mentions of specific geographic locations and establishments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Plasma

-0.75

asks

-0.69

clip

-0.68

SPD

-0.67

ape

-0.65

 Kers

-0.65

 plasma

-0.64

 gardening

-0.64

 Pasadena

-0.63

Henry

-0.62

POSITIVE LOGITS

Taj

4.54

 Uzbek

1.73

Ish

1.28

Kyr

1.25

 Tata

1.23

 Hilton

1.23

zbek

1.22

 Azerbai

1.17

 Monteneg

1.17

 Kazakh

1.16

Activations Density 0.030%

places or proper nouns related to specific countries and locations

mentions of specific geographic locations and establishments

No Comments

No Known Activations