INDEX

Explanations

words related to uncleanliness or unethical activities

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "dirty" 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 @ 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

yip

-0.97

*/(

-0.85

HCR

-0.83

ajor

-0.82

iphate

-0.81

nesota

-0.80

mare

-0.79

ommel

-0.79

icultural

-0.77

uther

-0.77

POSITIVE LOGITS

 laundry

1.03

 linen

0.93

 dirty

0.87

 cleaner

0.82

 diaper

0.79

 tricks

0.77

dirty

0.76

clean

0.76

 mole

0.76

 diapers

0.76

Activations Density 0.015%

words related to uncleanliness or unethical activities

references to the concept of "dirty" in various contexts

No Comments

No Known Activations