INDEX

Explanations

phrases related to negative sentiments or situations

oai_token-act-pair · gpt-3.5-turbo

negative phrases or concepts expressed in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Ventura

-0.77

 Dickinson

-0.73

 privately

-0.72

 Canter

-0.67

çīĪ

-0.66

 Shutterstock

-0.65

 Feder

-0.65

 Schumer

-0.64

 trough

-0.64

 Houth

-0.64

POSITIVE LOGITS

sounding

1.83

looking

1.70

enough

1.57

minded

1.41

yet

1.33

ness

1.26

colored

1.24

eyed

1.24

bodied

1.13

but

1.13

Activations Density 0.071%

phrases related to negative sentiments or situations

negative phrases or concepts expressed in the text

No Comments

No Known Activations