INDEX

Explanations

words related to items associated with dining or eating, specifically identifying napkins

oai_token-act-pair · gpt-3.5-turbo

references to napkins

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Mand

-0.75

²¾

-0.75

ourke

-0.73

leys

-0.72

Relations

-0.71

Chance

-0.68

ley

-0.68

RAY

-0.67

lev

-0.66

×Ļ×

-0.66

POSITIVE LOGITS

oleon

1.43

nap

1.30

puter

0.89

edia

0.86

olitan

0.79

anqu

0.79

olicy

0.78

ublic

0.78

eday

0.77

pav

0.75

Activations Density 0.011%

words related to items associated with dining or eating, specifically identifying napkins

references to napkins

No Comments

No Known Activations