INDEX

Explanations

descriptors of being the most recent or up-to-date

oai_token-act-pair · gpt-3.5-turbo

phrases related to recent events or updates

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

lying

-0.73

lain

-0.70

-0.69

Klu

-0.66

unction

-0.66

-0.64

erving

-0.63

-0.62

ammad

-0.61

POSITIVE LOGITS

 arrivals

0.82

 installment

0.79

 imaginable

0.72

 developments

0.68

 headlines

0.67

 trending

0.67

 refresh

0.65

Charges

0.64

DragonMagazine

0.63

 iteration

0.63

Activations Density 0.106%

descriptors of being the most recent or up-to-date

phrases related to recent events or updates

No Comments

No Known Activations