INDEX

Explanations

mentions of locations or entities related to Sweden

oai_token-act-pair · gpt-3.5-turbo

references to Sweden and Swedish culture

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Mayweather

-0.88

ebus

-0.71

ouston

-0.66

++++

-0.64

clusively

-0.62

orer

-0.62

ilage

-0.61

ciating

-0.61

 Indianapolis

-0.60

aneously

-0.60

POSITIVE LOGITS

borg

1.65

sson

1.20

Ã¥

1.05

lund

1.04

 Andersen

1.01

 Finland

1.01

 Sweden

1.00

holm

1.00

 Ragnar

0.98

 Kron

0.96

Activations Density 0.078%

mentions of locations or entities related to Sweden

references to Sweden and Swedish culture

No Comments

No Known Activations