INDEX

Explanations

phrases related to personal backgrounds and occupations

oai_token-act-pair · gpt-3.5-turbo

occurrences of various professional identifiers and roles in descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

zers

-0.77

ppings

-0.73

¬¼

-0.72

uz

-0.72

ullivan

-0.71

lash

-0.70

estones

-0.69

ights

-0.69

keyes

-0.69

isi

-0.68

POSITIVE LOGITS

 albeit

0.99

 whose

0.90

huh

0.86

0.85

 specializing

0.83

aka

0.83

 although

0.83

etc

0.80

 though

0.80

who

0.79

Activations Density 0.328%

phrases related to personal backgrounds and occupations

occurrences of various professional identifiers and roles in descriptions

No Comments

No Known Activations