INDEX

Explanations

words related to introducing or implementing something new or additional

oai_token-act-pair · gpt-3.5-turbo

references to new features or additions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Zip

-0.76

aminer

-0.75

ufact

-0.67

ulhu

-0.67

uddin

-0.65

ashtra

-0.65

acebook

-0.64

itta

-0.64

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

-0.63

phe

-0.62

POSITIVE LOGITS

bies

1.36

bie

1.31

 Zealand

1.07

foundland

1.01

 arrivals

0.96

 discoveries

0.93

 additions

0.92

 entrants

0.89

ãĤ´

0.89

found

0.83

Activations Density 0.109%

words related to introducing or implementing something new or additional

references to new features or additions

No Comments

No Known Activations