INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

references to royalty or royal entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ither

-0.72

perature

-0.66

ostics

-0.65

ofi

-0.64

oko

-0.61

mitted

-0.61

 stains

-0.61

 msec

-0.60

aping

-0.60

iland

-0.60

POSITIVE LOGITS

Royal

1.33

 manpower

0.69

ellation

0.68

isance

0.67

FSA

0.64

eele

0.63

AAA

0.62

tyard

0.61

sterdam

0.61

CBO

0.60

Activations Density 0.003%

proper nouns

references to royalty or royal entities

No Comments

No Known Activations