INDEX

Explanations

positive descriptions of food and cooking experiences

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

expressions of personal experiences and recommendations related to products or food

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

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_post

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

cknow

-0.80

actionDate

-0.72

»Ĵ

-0.70

 initiation

-0.65

Pause

-0.65

 pestic

-0.64

ItemThumbnailImage

-0.64

farious

-0.64

DragonMagazine

-0.63

lihood

-0.63

POSITIVE LOGITS

 tasted

1.13

 smelled

1.09

 worked

1.04

 tastes

1.04

 didnt

1.03

 feels

1.02

 smells

1.01

 looked

1.01

 reminds

1.00

 sounded

0.99

Activations Density 0.238%

positive descriptions of food and cooking experiences

expressions of personal experiences and recommendations related to products or food

No Comments

No Known Activations