INDEX

Explanations

locations where people have lived or are currently living

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the duration or history of someone's experiences and relationships

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

âĺħ

-0.76

 Sabha

-0.75

article

-0.66

Pierre

-0.65

iru

-0.64

MJ

-0.64

ember

-0.63

ophe

-0.62

MG

-0.62

Rap

-0.62

POSITIVE LOGITS

 since

0.95

 repeatedly

0.90

 awhile

0.84

 countless

0.84

 extensively

0.83

 lately

0.77

 previously

0.77

 unsuccessfully

0.76

 thoroughly

0.74

 plenty

0.74

Activations Density 0.415%

locations where people have lived or are currently living

phrases indicating the duration or history of someone's experiences and relationships

No Comments

No Known Activations