INDEX

Explanations

adjectives describing the quality of something, especially contrasting good and bad

oai_token-act-pair · gpt-3.5-turbo

concepts of good versus bad, particularly in a comparative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

atus

-0.73

mith

-0.73

ãĥĹ

-0.72

yth

-0.70

forts

-0.67

imum

-0.65

gif

-0.64

icum

-0.64

ãĤ¸

-0.63

 Williamson

-0.63

POSITIVE LOGITS

 alike

1.12

 depending

1.00

 sides

0.78

 respectively

0.77

 stripes

0.74

depending

0.73

 sectors

0.73

 versions

0.71

 coasts

0.66

 wings

0.65

Activations Density 0.240%

adjectives describing the quality of something, especially contrasting good and bad

concepts of good versus bad, particularly in a comparative context

No Comments

No Known Activations