INDEX

Explanations

phrases related to dialogue and interactions between characters

oai_token-act-pair · gpt-3.5-turbo

punctuation marks that express surprise or emphasis

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

irable

-0.86

regate

-0.78

rss

-0.75

osate

-0.74

iat

-0.71

rahim

-0.71

=(

-0.70

 Tycoon

-0.69

isi

-0.68

azeera

-0.67

POSITIVE LOGITS

 grinning

1.24

 nodding

1.23

 clutching

1.21

 blinking

1.20

 eyeing

1.17

 waving

1.15

 leaning

1.14

 smiling

1.12

 staring

1.12

 noticing

1.12

Activations Density 0.150%

phrases related to dialogue and interactions between characters

punctuation marks that express surprise or emphasis

No Comments

No Known Activations