INDEX

Explanations

personalities or professions represented by proper nouns

oai_token-act-pair · gpt-3.5-turbo

titles and roles associated with various professions and expertise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

etheless

-0.80

 conclud

-0.77

 compr

-0.75

atever

-0.72

arching

-0.68

âķĲâķĲ

-0.68

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.67

ilogy

-0.66

WARNING

-0.66

×©

-0.63

POSITIVE LOGITS

 extraord

1.55

 Mike

1.17

 Joey

1.15

 Jason

1.14

 Kevin

1.13

 Cody

1.11

 Dave

1.11

 Randy

1.10

 Andy

1.10

 Steve

1.10

Activations Density 0.230%

personalities or professions represented by proper nouns

titles and roles associated with various professions and expertise

No Comments

No Known Activations