INDEX

Explanations

words related to personal qualities or descriptions of people

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals and their characteristics or actions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ramid

-0.83

sequent

-0.73

encies

-0.71

ð

-0.70

iners

-0.69

inav

-0.68

uates

-0.67

ä½¿

-0.66

negie

-0.66

rued

-0.65

POSITIVE LOGITS

 brilliant

1.03

 fearless

1.03

 hilarious

1.01

 charismatic

0.99

 charming

0.99

 adorable

0.99

 terrific

0.97

 handsome

0.96

 gorgeous

0.95

 fantastic

0.94

Activations Density 0.233%

words related to personal qualities or descriptions of people

references to specific individuals and their characteristics or actions

No Comments

No Known Activations