INDEX

Explanations

punctuation marks, specifically periods and commas

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

titles or labels represented by words ending with a period and followed by numbers within a text document

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 lifes

-0.84

 pract

-0.80

 occupations

-0.79

 actu

-0.79

 decriminal

-0.78

 reflex

-0.77

 clerks

-0.77

 statistical

-0.77

 mosqu

-0.76

 disadvantaged

-0.74

POSITIVE LOGITS

advertising

1.19

 Speaking

1.17

Fans

1.16

Speaking

1.12

Rum

1.10

 Interestingly

1.09

 Previously

1.05

Previously

1.05

 Regarding

1.03

 Additionally

1.03

Activations Density 0.496%

punctuation marks, specifically periods and commas

titles or labels represented by words ending with a period and followed by numbers within a text document

No Comments

No Known Activations