INDEX

Explanations

words related to electricity or electric-powered items

oai_token-act-pair · gpt-3.5-turbo

references to electric or electronic technology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 McKenna

-0.77

 Haku

-0.71

Beck

-0.71

00200000

-0.70

allery

-0.70

 Cull

-0.69

 Michaels

-0.69

MY

-0.68

xual

-0.66

ort

-0.66

POSITIVE LOGITS

 propulsion

0.91

 electric

0.89

 motors

0.88

 Powered

0.87

 currents

0.86

 heater

0.85

 grids

0.82

 powered

0.81

 appliances

0.81

onda

0.80

Activations Density 0.009%

words related to electricity or electric-powered items

references to electric or electronic technology

No Comments

No Known Activations