INDEX

Explanations

mentions of the abbreviation 'ia' coupled with a high activation value

oai_token-act-pair · gpt-3.5-turbo

the presence of specific character or name fragments throughout the document

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

inances

-0.79

alore

-0.79

mans

-0.75

alogue

-0.74

soDeliveryDate

-0.74

 glim

-0.73

iple

-0.72

ortment

-0.69

ergy

-0.68

ager

-0.67

POSITIVE LOGITS

zza

1.35

ppe

1.17

pered

1.00

BILITY

0.96

ccess

0.95

 nerv

0.94

 coli

0.93

ffe

0.93

ppa

0.93

ÅĤ

0.91

Activations Density 0.035%

mentions of the abbreviation 'ia' coupled with a high activation value

the presence of specific character or name fragments throughout the document

No Comments

No Known Activations