INDEX

Explanations

ingredients or cooking-related words in recipes

oai_token-act-pair · gpt-3.5-turbo

references to specific items or recipes

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

nings

-0.78

arson

-0.74

andum

-0.73

witz

-0.71

isms

-0.71

occup

-0.69

adata

-0.67

ussen

-0.67

lev

-0.66

anamo

-0.66

POSITIVE LOGITS

 lovely

0.91

guy

0.86

 lowly

0.85

 amazing

0.85

 nifty

0.83

 particular

0.83

 gorgeous

0.83

 handy

0.79

 sucker

0.78

 adorable

0.78

Activations Density 0.219%

ingredients or cooking-related words in recipes

references to specific items or recipes

No Comments

No Known Activations