INDEX

Explanations

mentions of different size categories like small, medium, or large

oai_token-act-pair · gpt-3.5-turbo

phrases that specify levels of size or scale, particularly in relation to business or operations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

eries

-0.83

 VIDEOS

-0.70

edia

-0.67

 fingerprints

-0.67

 Airl

-0.66

 Alive

-0.66

Images

-0.65

 Tears

-0.64

ModLoader

-0.64

ETS

-0.64

POSITIVE LOGITS

 intermediate

1.26

 mids

1.07

mid

1.05

 middle

1.01

 upper

0.99

 tert

0.97

 midrange

0.95

 medium

0.92

 advanced

0.91

 late

0.89

Activations Density 0.109%

mentions of different size categories like small, medium, or large

phrases that specify levels of size or scale, particularly in relation to business or operations

No Comments

No Known Activations