INDEX

Explanations

technical terms related to computer chips

oai_token-act-pair · gpt-3.5-turbo

references to chips and chip technology

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

æĢ

-0.72

 Stranger

-0.71

Vaj

-0.65

 Dayton

-0.64

etheless

-0.63

 gearing

-0.63

 Scandinav

-0.63

 Kamp

-0.63

lihood

-0.63

 Graves

-0.62

POSITIVE LOGITS

otle

1.43

 chip

0.99

 chips

0.96

washer

0.96

chip

0.94

mun

0.90

 implanted

0.86

board

0.86

intosh

0.85

Chip

0.83

Activations Density 0.009%

technical terms related to computer chips

references to chips and chip technology

No Comments

No Known Activations