INDEX

Explanations

comments or opinions expressed in a text

oai_token-act-pair · gpt-3.5-turbo

references to comments and interactions in a discussion or article

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

arius

-0.75

abil

-0.75

ens

-0.69

hound

-0.68

iversal

-0.68

 breaker

-0.65

 behold

-0.63

pired

-0.62

within

-0.62

vantage

-0.62

POSITIVE LOGITS

 comments

3.69

 remarks

2.71

 comment

2.48

 Comments

2.43

comments

2.05

 commenters

1.96

 commenting

1.96

Comments

1.83

 replies

1.81

 remark

1.78

Activations Density 0.023%

comments or opinions expressed in a text

references to comments and interactions in a discussion or article

No Comments

No Known Activations