INDEX

Explanations

specific entities such as institutions, locations, and organizations

oai_token-act-pair · gpt-3.5-turbo

names of institutions, places, or significant entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 oneself

-0.96

 extrad

-0.71

theless

-0.65

etheless

-0.65

 sender

-0.63

 yourself

-0.63

 deaf

-0.63

 muted

-0.63

 Hond

-0.63

 USPS

-0.62

POSITIVE LOGITS

ometown

0.83

 Address

0.79

 Advisory

0.79

Wallet

0.77

 Crimes

0.75

 Series

0.75

Series

0.73

ertation

0.72

 Strategy

0.72

irlfriend

0.72

Activations Density 0.323%

specific entities such as institutions, locations, and organizations

names of institutions, places, or significant entities

No Comments

No Known Activations