INDEX

Explanations

references to genomic processes or terms

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

 loose

-1.66

ĥ½

-1.64

'</

-1.63

athing

-1.61

³

-1.60

č↵

-1.59

)$$

-1.56

');

-1.55

>';

-1.53

¼

-1.53

POSITIVE LOGITS

iliary

1.59

idian

1.58

etica

1.58

fecture

1.55

vill

1.53

fonts

1.51

zilla

1.50

 arsenal

1.48

coin

1.46

assic

1.46

Activations Density 0.059%

references to genomic processes or terms

No Comments

No Known Activations