INDEX

Explanations

references to additional or bonus content, such as special features or extras

oai_token-act-pair · gpt-3.5-turbo

references to additional content or features associated with media or products

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

izont

-0.98

²¾

-0.79

iard

-0.75

ggle

-0.74

£ı

-0.72

 Weed

-0.70

berman

-0.68

SHIP

-0.67

cko

-0.67

Īè

-0.65

POSITIVE LOGITS

ordinary

0.99

 extras

0.83

members

0.74

asio

0.72

ilib

0.72

ordinate

0.70

ortal

0.70

member

0.70

asks

0.69

visor

0.69

Activations Density 0.015%

references to additional or bonus content, such as special features or extras

references to additional content or features associated with media or products

No Comments

No Known Activations