INDEX

Explanations

punctuation marks

oai_token-act-pair · gpt-3.5-turbo

punctuation, specifically commas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Loading

-0.54

ONSORED

-0.52

abee

-0.52

 guiActiveUn

-0.50

atern

-0.48

 Pipeline

-0.47

 referen

-0.46

ãĤ´ãĥ³

-0.46

¬¼

-0.45

iola

-0.45

POSITIVE LOGITS

huh

1.10

 please

0.96

etc

0.90

 albeit

0.88

but

0.84

 namely

0.84

 preferably

0.83

eh

0.81

 haha

0.79

please

0.79

Activations Density 0.426%

punctuation marks

punctuation, specifically commas

No Comments

No Known Activations