INDEX

Explanations

names of individuals and organizations

oai_token-act-pair · gpt-3.5-turbo

proper nouns or specific identifiers

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Cherokee

-0.91

 coral

-0.86

 Cyrus

-0.86

 Scorpion

-0.85

 Manila

-0.85

 Scorp

-0.81

 Indian

-0.79

 snakes

-0.79

 lime

-0.79

 Coral

-0.78

POSITIVE LOGITS

Ã¼

1.64

Ã¶

1.61

Ã¤

1.55

Ã¼r

1.54

Ã¶r

1.43

ortmund

1.36

Ã¶n

1.36

ÃŁ

1.34

 Bundes

1.25

Germany

1.23

Activations Density 0.148%

names of individuals and organizations

proper nouns or specific identifiers

No Comments

No Known Activations