INDEX

Explanations

sequential instructions or steps provided in text format

oai_token-act-pair · gpt-3.5-turbo

phrases that suggest actionable steps or recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Blog

-0.48

canon

-0.46

 Cannes

-0.45

SPI

-0.44

 Lich

-0.42

 Ladies

-0.42

 Editor

-0.42

 Herald

-0.41

 Anat

-0.41

 Lets

-0.41

POSITIVE LOGITS

aneously

0.55

usercontent

0.54

.''.

0.50

 contrace

0.49

soever

0.49

 outwe

0.49

 (âĪĴ

0.48

 themselves

0.48

'."

0.48

 deductions

0.47

Activations Density 2.100%

sequential instructions or steps provided in text format

phrases that suggest actionable steps or recommendations

No Comments

No Known Activations