INDEX

Explanations

exclamations indicating strong emotions or opinions

oai_token-act-pair · gpt-3.5-turbo

exclamatory statements expressing strong emotions or reactions

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

undai

-0.75

 conduc

-0.74

 destro

-0.70

 encount

-0.69

rament

-0.68

 corrid

-0.68

itates

-0.67

yrus

-0.66

 glim

-0.66

itated

-0.66

POSITIVE LOGITS

?!

1.43

@#&

1.30

@#

1.23

#$

1.23

 Surely

1.08

 Anyway

1.06

ãĢį

1.06

:-)

1.05

Isn

1.04

 ðŁĺ

1.03

Activations Density 0.071%

exclamations indicating strong emotions or opinions

exclamatory statements expressing strong emotions or reactions

No Comments

No Known Activations