INDEX

Explanations

texts related to specific data formatting or structured text

oai_token-act-pair · gpt-3.5-turbo

occurrences of slashes or divisions in text

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

 Beir

-0.86

terday

-0.72

acters

-0.72

abies

-0.69

 defe

-0.67

etime

-0.66

idation

-0.64

 explan

-0.64

onic

-0.63

uration

-0.62

POSITIVE LOGITS

ËĪ

1.54

usr

1.02

wcsstore

0.84

etc

0.83

Via

0.82

Film

0.82

home

0.80

 Shutterstock

0.79

join

0.75

tg

0.75

Activations Density 0.026%

texts related to specific data formatting or structured text

occurrences of slashes or divisions in text

No Comments

No Known Activations