INDEX

Explanations

phrases related to scientific theories

oai_token-act-pair · gpt-3.5-turbo

references to various theories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 ---------

-0.67

awn

-0.65

Serv

-0.63

Lic

-0.61

 unwanted

-0.61

 Borders

-0.61

elcome

-0.60

Host

-0.60

Contact

-0.58

lete

-0.57

POSITIVE LOGITS

 theory

3.84

 Theory

2.85

 theories

2.65

 theorists

2.39

 hypothesis

2.21

 theorist

2.17

 theoret

1.79

 theor

1.75

 hypotheses

1.62

 theoretical

1.54

Activations Density 0.023%

phrases related to scientific theories

references to various theories

No Comments

No Known Activations