INDEX

Explanations

recognizable names or phrases related to famous personalities, places, or events

oai_token-act-pair · gpt-3.5-turbo

references to specific dates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 dime

-0.73

Reviewer

-0.72

 thirds

-0.67

 personally

-0.58

 independents

-0.57

LLOW

-0.56

 retard

-0.55

 disabilities

-0.54

RO

-0.53

tru

-0.52

POSITIVE LOGITS

vier

1.32

itor

1.26

esville

1.25

itors

1.16

ice

1.12

eway

1.03

elle

0.94

imus

0.91

umping

0.89

umper

0.87

Activations Density 0.022%

recognizable names or phrases related to famous personalities, places, or events

references to specific dates

No Comments

No Known Activations