INDEX

Explanations

phrases indicating communication or interaction with others

oai_token-act-pair · gpt-3.5-turbo

expressions of welcoming and community engagement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

stroke

-0.63

FU

-0.61

 Ambro

-0.60

 imaginable

-0.59

pex

-0.58

emer

-0.58

\<

-0.58

thinkable

-0.57

éĹ

-0.57

 syndrome

-0.57

POSITIVE LOGITS

 ourselves

1.31

 ours

0.84

our

0.80

ngth

0.74

yss

0.71

oday

0.71

psons

0.66

 parted

0.65

 hereby

0.64

mble

0.64

Activations Density 0.911%

phrases indicating communication or interaction with others

expressions of welcoming and community engagement

No Comments

No Known Activations