INDEX

Explanations

mentions of the word "sacred"

oai_token-act-pair · gpt-3.5-turbo

references to sacredness or sacrificial themes

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

ITNESS

-0.77

 Cheong

-0.76

 Turk

-0.66

NING

-0.65

ITH

-0.65

 Imran

-0.64

 expectancy

-0.64

agher

-0.64

è¦ļéĨĴ

-0.63

DERR

-0.63

POSITIVE LOGITS

ificial

1.29

rament

1.19

char

1.12

rum

1.05

anus

1.04

het

1.01

raf

1.00

ram

0.97

ros

0.95

ral

0.95

Activations Density 0.015%

mentions of the word "sacred"

references to sacredness or sacrificial themes

No Comments

No Known Activations