INDEX

Explanations

questions ending in a question mark

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and rhetorical questions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tremend

-0.90

 millenn

-0.86

 carbohyd

-0.82

¥ŀ

-0.82

 furnace

-0.81

 advoc

-0.78

 destro

-0.77

ultz

-0.76

iann

-0.75

 concess

-0.73

POSITIVE LOGITS

 Anyway

1.91

Anyway

1.60

;)

1.37

 ðŁĻĤ

1.34

But

1.31

:)

1.27

:-)

1.26

 Especially

1.26

 Nonetheless

1.23

 Nevertheless

1.19

Activations Density 0.447%

questions ending in a question mark

punctuation marks and rhetorical questions

No Comments

No Known Activations