INDEX

Explanations

words related to substances or activities like drinking or drugs

oai_token-act-pair · gpt-3.5-turbo

references to drinking habits or beverages

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

-0.88

eers

-0.86

SHIP

-0.78

quo

-0.73

eering

-0.68

 Leilan

-0.66

eer

-0.63

hire

-0.62

 Pose

-0.61

 Economic

-0.59

POSITIVE LOGITS

inking

1.34

agons

1.30

inks

1.24

unk

1.21

ink

1.16

ags

1.15

ifting

1.14

agging

1.14

ifts

1.13

ifter

1.13

Activations Density 0.026%

words related to substances or activities like drinking or drugs

references to drinking habits or beverages

No Comments

No Known Activations