INDEX

Explanations

technology-related terms and entities, like companies, positions, and products

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to people, places, and organizations

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

ertodd

-0.66

uana

-0.63

 Swordsman

-0.62

arthed

-0.57

ificent

-0.56

ITNESS

-0.56

soDeliveryDate

-0.55

 Marketable

-0.55

thumbnails

-0.54

 Cancel

-0.53

POSITIVE LOGITS

 shores

0.59

 eligibility

0.57

 sensibilities

0.56

 efforts

0.56

goers

0.55

 itch

0.54

ans

0.54

 fortunes

0.54

ankind

0.54

ism

0.53

Activations Density 0.639%

technology-related terms and entities, like companies, positions, and products

proper nouns related to people, places, and organizations

No Comments

No Known Activations