INDEX

Explanations

references to rules, regulations, or obligations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating obligations or mandates

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

 Ãľ

-0.59

izoph

-0.57

Alz

-0.55

ultz

-0.53

fal

-0.52

ovember

-0.52

 Mong

-0.51

 Valkyrie

-0.50

FK

-0.49

Rez

-0.48

POSITIVE LOGITS

to

1.00

by

0.94

ICLE

0.74

 aback

0.73

BY

0.73

pta

0.71

TO

0.71

by

0.71

to

0.69

 annually

0.66

Activations Density 0.129%

references to rules, regulations, or obligations

phrases indicating obligations or mandates

No Comments

No Known Activations