INDEX

Explanations

food-related terms and recipes

oai_token-act-pair · gpt-3.5-turbo

elements related to cooking and 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

 Samar

-0.77

 estranged

-0.75

 interpre

-0.71

 forthcoming

-0.71

 constitutionally

-0.71

 affirmative

-0.70

 interven

-0.70

 orchestr

-0.69

 undocumented

-0.68

 extradition

-0.68

POSITIVE LOGITS

Ingredients

1.28

Recipe

1.20

Instruct

0.99

Step

0.98

Want

0.95

Rated

0.95

Nut

0.94

Heat

0.92

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.92

................................................................

0.92

Activations Density 0.298%

food-related terms and recipes

elements related to cooking and recipes

No Comments

No Known Activations