INDEX

Explanations

terms indicating sequencing or order

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/2-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.2.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

¿½

-1.62

 progress

-1.55

 CONTRIBUTORS

-1.52

 continued

-1.49

 progresses

-1.44

 caution

-1.43

 Progress

-1.42

ths

-1.42

'?"

-1.41

 rush

-1.41

POSITIVE LOGITS

ache

2.35

abase

2.07

illion

1.79

aches

1.74

ificate

1.64

ariate

1.60

arda

1.53

ape

1.50

otec

1.49

alian

1.49

Activations Density 0.535%

terms indicating sequencing or order

No Comments

No Known Activations