INDEX

Explanations

phrases related to happiness and satisfaction

oai_token-act-pair · gpt-3.5-turbo

expressions of happiness and positive sentiments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

•layer 5 - happiness

No Comments

Negative Logits

isSpecialOrderable

-0.82

soDeliveryDate

-0.81

Wan

-0.76

ngth

-0.72

 alters

-0.68

hill

-0.64

aredevil

-0.61

å£«

-0.60

orno

-0.60

ware

-0.60

POSITIVE LOGITS

whel

0.68

clus

0.63

cel

0.62

 applaud

0.62

rix

0.61

 Frog

0.61

ovy

0.60

iola

0.60

idays

0.59

 Ribbon

0.59

Activations Density 0.136%

phrases related to happiness and satisfaction

expressions of happiness and positive sentiments

No Comments

No Known Activations