INDEX

Explanations

phrases related to physical locations or expanses

oai_token-act-pair · gpt-3.5-turbo

terms related to environmental or spatial impact

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

erest

-0.73

icum

-0.69

actionDate

-0.68

acha

-0.68

regulated

-0.67

ippery

-0.65

imar

-0.64

ending

-0.63

 antidepressants

-0.63

ifying

-0.63

POSITIVE LOGITS

 footprint

1.45

 footprints

1.09

Þ

0.72

sheet

0.68

 perimeter

0.66

sheets

0.65

 imprint

0.65

 Takeru

0.64

layer

0.64

 silhouette

0.64

Activations Density 0.012%

phrases related to physical locations or expanses

terms related to environmental or spatial impact

No Comments

No Known Activations