INDEX

Explanations

periods at the end of sentences

oai_token-act-pair · gpt-3.5-turbo

sentences or phrases that potentially indicate the conclusion of a thought or section

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

 reversible

-0.85

 unpredict

-0.80

 desper

-0.78

 deflation

-0.78

 sideways

-0.77

 slightest

-0.77

 thinner

-0.77

 unexplained

-0.76

 incon

-0.76

 upgr

-0.75

POSITIVE LOGITS

Its

1.35

 Called

1.27

 Participants

1.27

 Their

1.26

 Organizations

1.24

 Volunteers

1.23

 Together

1.17

 Among

1.17

 Members

1.16

 Funding

1.16

Activations Density 0.394%

periods at the end of sentences

sentences or phrases that potentially indicate the conclusion of a thought or section

No Comments

No Known Activations