INDEX

Explanations

years represented as numbers, such as "30-year" or "60-minute"

oai_token-act-pair · gpt-3.5-turbo

numerical data related to time spans and measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ModLoader

-0.85

ophen

-0.72

 Contents

-0.70

 è£ıè

-0.70

 Flavoring

-0.69

fml

-0.64

áµ

-0.63

ificantly

-0.61

 contiguous

-0.59

culosis

-0.59

POSITIVE LOGITS

eger

0.75

phia

0.72

quez

0.70

oak

0.69

kens

0.68

itte

0.66

oche

0.66

veland

0.63

ANN

0.63

deg

0.62

Activations Density 0.272%

years represented as numbers, such as "30-year" or "60-minute"

numerical data related to time spans and measurements

No Comments

No Known Activations