INDEX

Explanations

phrases or contexts where something is logically sound or reasonable

oai_token-act-pair · gpt-3.5-turbo

the phrase "makes sense" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aunting

-0.73

psy

-0.72

 helicop

-0.66

 repaired

-0.65

 downed

-0.65

onder

-0.65

 succumbed

-0.63

 toured

-0.63

ritical

-0.63

enne

-0.62

POSITIVE LOGITS

hift

0.79

SourceFile

0.75

 Marketable

0.74

GROUP

0.73

DragonMagazine

0.73

imaru

0.72

CLASSIFIED

0.72

FontSize

0.71

ãħĭ

0.70

soDeliveryDate

0.70

Activations Density 0.014%

phrases or contexts where something is logically sound or reasonable

the phrase "makes sense" in various contexts

No Comments

No Known Activations