INDEX

Explanations

phrases related to noteworthy or impactful occurrences

oai_token-act-pair · gpt-3.5-turbo

instances of the word "phenomenon" and its variants, indicating discussions of notable occurrences or trends

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

gur

-0.96

zar

-0.74

onz

-0.74

bey

-0.73

oglu

-0.71

zik

-0.69

lems

-0.69

zer

-0.67

pees

-0.67

llah

-0.66

POSITIVE LOGITS

 phenomenon

0.99

ually

0.88

ively

0.87

 phenomena

0.85

ĸļ

0.83

ional

0.82

 occurring

0.79

 whereby

0.79

rences

0.78

uate

0.76

Activations Density 0.015%

phrases related to noteworthy or impactful occurrences

instances of the word "phenomenon" and its variants, indicating discussions of notable occurrences or trends

No Comments

No Known Activations