INDEX

Explanations

phrases related to praising or celebrating

oai_token-act-pair · gpt-3.5-turbo

terms that express acclaim or recognition

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

Not in Any Lists

No Comments

Negative Logits

ascript

-0.81

 diapers

-0.79

omaly

-0.78

uyomi

-0.77

hing

-0.77

thia

-0.76

 diaper

-0.71

 captcha

-0.69

 inhib

-0.68

 stomach

-0.68

POSITIVE LOGITS

 praises

0.81

 virtues

0.78

enance

0.77

Supporters

0.73

 Friedman

0.71

 hailed

0.70

Stand

0.70

GROUND

0.66

 loudly

0.66

 acclaim

0.65

Activations Density 0.041%

phrases related to praising or celebrating

terms that express acclaim or recognition

No Comments

No Known Activations