INDEX

Explanations

phrases indicating completeness or totality

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulas

-1.51

eeper

-1.47

chers

-1.46

ierno

-1.45

 Department

-1.44

®

-1.43

icki

-1.41

ember

-1.39

 Method

-1.38

ĨĴ

-1.38

POSITIVE LOGITS

 blown

1.87

body

1.66

blown

1.60

 onto

1.50

court

1.49

street

1.48

metal

1.46

filling

1.45

bod

1.43

wave

1.43

Activations Density 0.354%

phrases indicating completeness or totality

No Comments

No Known Activations