INDEX

Explanations

mentions of different nationalities or origins, specifically focusing on Brazilian and Argentine individuals

oai_token-act-pair · gpt-3.5-turbo

references to various nationalities and ethnicities of football players

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

illance

-0.86

fman

-0.85

igslist

-0.84

regon

-0.79

lease

-0.79

utterstock

-0.78

paio

-0.78

 Hobby

-0.76

illary

-0.75

rooms

-0.75

POSITIVE LOGITS

 youngster

1.06

 footballer

1.03

 winger

1.01

 midfielder

0.99

 goalkeeper

0.94

 striker

0.94

duo

0.94

 defender

0.92

 superstar

0.90

 Football

0.88

Activations Density 0.068%

mentions of different nationalities or origins, specifically focusing on Brazilian and Argentine individuals

references to various nationalities and ethnicities of football players

No Comments

No Known Activations