INDEX

Explanations

terms related to data management and protection

oai_token-act-pair · gpt-3.5-turbo

phrases related to large-scale issues and actions needed to address them

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

\-

-0.63

Kids

-0.61

 âĨĳ

-0.56

 Prayer

-0.55

twitter

-0.55

inar

-0.55

issors

-0.54

aldo

-0.53

laugh

-0.53

vice

-0.53

POSITIVE LOGITS

 these

1.29

 such

1.28

 this

1.08

 truly

0.98

 those

0.96

these

0.93

 meaningful

0.92

 genuinely

0.90

 genuine

0.86

 them

0.85

Activations Density 0.705%

terms related to data management and protection

phrases related to large-scale issues and actions needed to address them

No Comments

No Known Activations