INDEX

Explanations

instances of "elimination" in various contexts

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

ĥ½

-1.99

ĨĴ

-1.82

-1.79

ness

-1.68

romycin

-1.63

*=

-1.54

½

-1.52

ships

-1.48

rael

-1.44

¿½

-1.41

POSITIVE LOGITS

 gaps

1.82

 compartments

1.67

naire

1.66

 slots

1.59

esis

1.53

 compartment

1.52

 circuits

1.46

 chambers

1.46

 momento

1.46

ogue

1.46

Activations Density 0.055%

instances of "elimination" in various contexts

No Comments

No Known Activations