INDEX

Explanations

lines of code or structured data

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

ĥ½

-1.93

 anonymity

-1.66

 others

-1.65

 nominal

-1.59

 angels

-1.53

imony

-1.51

 anyone

-1.50

 someone

-1.48

rians

-1.48

uties

-1.46

POSITIVE LOGITS

ARRAY

1.90

iewicz

1.90

wheel

1.84

works

1.72

paste

1.70

umber

1.64

hole

1.61

orest

1.59

STRING

1.59

brew

1.57

Activations Density 0.009%

lines of code or structured data

No Comments

No Known Activations