INDEX

Explanations

phrases related to a kitchen or kitchen-related activities

oai_token-act-pair · gpt-3.5-turbo

references to kitchens

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

âĢ¢âĢ¢

-0.78

 Trooper

-0.71

ullah

-0.70

 USAF

-0.69

gow

-0.67

vernment

-0.67

idas

-0.65

worthiness

-0.65

utic

-0.64

RAY

-0.64

POSITIVE LOGITS

washer

1.00

 cabinets

0.93

 kitchen

0.91

 kitchens

0.90

 refrigerator

0.86

 cooks

0.85

tops

0.84

 sink

0.84

nut

0.84

ware

0.82

Activations Density 0.012%

phrases related to a kitchen or kitchen-related activities

references to kitchens

No Comments

No Known Activations