INDEX

Explanations

mentions of the word 'ordinary'

oai_token-act-pair · gpt-3.5-turbo

references to the concept of ordinary people or ordinary life

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

olic

-0.73

rogram

-0.71

asus

-0.69

haw

-0.69

arth

-0.68

zik

-0.67

lez

-0.64

utra

-0.64

CVE

-0.63

AIDS

-0.63

POSITIVE LOGITS

ordinary

1.06

 ordinary

0.88

 mortals

0.87

 sized

0.87

isation

0.78

 everyday

0.76

cy

0.76

ised

0.76

 weekday

0.75

arily

0.74

Activations Density 0.015%

mentions of the word 'ordinary'

references to the concept of ordinary people or ordinary life

No Comments

No Known Activations