INDEX

Explanations

instances of "rules" and their implications

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 please

-1.72

afers

-1.68

![

-1.47

Âł↵

-1.44

rae

-1.41

days

-1.34

 waived

-1.34

lings

-1.34

 starter

-1.31

 soit

-1.30

POSITIVE LOGITS

ĨĴ

3.23

ĻĤ

2.79

Ń

2.63

£

2.48

ĸ

2.46

į

2.46

¾

2.44

ı

2.35

¥

2.31

º

2.30

Activations Density 0.007%

instances of "rules" and their implications

No Comments

No Known Activations