INDEX

Explanations

textual artifacts

oai_token-act-pair · gpt-3.5-turbo

phrases related to cultural dynamics or appropriation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 booster

-0.70

 consolidation

-0.69

 horizont

-0.67

 Minotaur

-0.66

 juggling

-0.66

assetsadobe

-0.66

 monop

-0.66

lain

-0.66

 mattress

-0.65

 planner

-0.65

POSITIVE LOGITS

Â

1.50

Â»

1.33

Â±

1.33

Â«

1.32

Â£

1.30

Â¢

1.25

\/

1.24

Â´

1.23

1.22

Â°

1.15

Activations Density 0.086%

textual artifacts

phrases related to cultural dynamics or appropriation

No Comments

No Known Activations