INDEX

Explanations

terms indicating majority or dominance

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

Ķ

-2.46

§

-2.34

Ĥ¬

-2.29

Ī

-2.23

»¿

-2.22

ĨĴ

-2.21

®

-2.12

ĵ

-2.09

«

-1.98

ĭ

-1.97

POSITIVE LOGITS

 heterosexual

1.46

 speaking

1.44

 enough

1.37

ships

1.37

 Justice

1.35

Than

1.31

 Murder

1.31

 suffice

1.31

 Yugoslav

1.31

 Words

1.30

Activations Density 0.009%

terms indicating majority or dominance

No Comments

No Known Activations