INDEX

Explanations

references to performers or performance-related terms

oai_token-act-pair · gpt-3.5-turbo

references to performers in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ocide

-0.68

ories

-0.67

ohyd

-0.63

£ı

-0.62

erald

-0.62

alg

-0.59

 Sett

-0.59

lly

-0.59

 Balance

-0.58

rain

-0.58

POSITIVE LOGITS

 performers

1.22

 performer

1.15

glers

0.93

goers

0.87

Artist

0.85

 exhib

0.82

OGR

0.82

 contestant

0.80

 contestants

0.80

 guiActiveUn

0.79

Activations Density 0.009%

references to performers or performance-related terms

references to performers in various contexts

No Comments

No Known Activations