INDEX

Explanations

phrases indicating direct speech

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, specifically commas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

vable

-0.79

rawdownloadcloneembedreportprint

-0.71

ords

-0.70

olves

-0.69

eland

-0.69

iciary

-0.68

 Flavoring

-0.66

rica

-0.66

rav

-0.66

imore

-0.66

POSITIVE LOGITS

hey

1.39

1.35

"'

1.32

'[

1.31

1.27

oh

1.19

1.17

\"

1.11

''

1.10

wow

1.10

Activations Density 0.094%

phrases indicating direct speech

punctuation marks, specifically commas

No Comments

No Known Activations