INDEX

Explanations

references to the word "Brow" or related 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

 late

-1.56

inib

-1.55

upon

-1.50

inct

-1.49

usterity

-1.48

 completeness

-1.45

aneously

-1.43

 great

-1.41

Į

-1.39

 nude

-1.39

POSITIVE LOGITS

ettes

2.00

esan

1.86

issance

1.85

ardo

1.73

ship

1.69

yards

1.67

caster

1.58

ouin

1.54

 Engineering

1.49

die

1.49

Activations Density 0.019%

references to the word "Brow" or related terms

No Comments

No Known Activations