INDEX

Explanations

words related to qualities or characteristics

oai_token-act-pair · gpt-3.5-turbo

words related to emotional states or conditions

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

ãĥĥãĥī

-0.65

bern

-0.64

ODE

-0.63

ellar

-0.62

verbs

-0.61

 Auschwitz

-0.61

roit

-0.60

ANN

-0.59

OY

-0.59

amen

-0.59

POSITIVE LOGITS

iness

1.16

terness

0.99

ness

0.97

nesses

0.94

ionage

0.81

 Flavoring

0.79

yy

0.77

liness

0.77

hip

0.75

osuke

0.75

Activations Density 0.028%

words related to qualities or characteristics

words related to emotional states or conditions

No Comments

No Known Activations