INDEX

Explanations

editions or versions of something

oai_token-act-pair · gpt-3.5-turbo

references to various editions of products or works

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

alyst

-0.83

irling

-0.77

estones

-0.73

abases

-0.72

abase

-0.71

abee

-0.70

ahon

-0.69

rogen

-0.69

ogram

-0.69

abolic

-0.68

POSITIVE LOGITS

gur

0.95

pires

0.79

quickShipAvailable

0.78

sense

0.72

BLE

0.67

ign

0.67

pas

0.65

hog

0.65

PRES

0.62

wards

0.62

Activations Density 0.054%

editions or versions of something

references to various editions of products or works

No Comments

No Known Activations