INDEX

Explanations

definitions or meanings of 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

§

-1.83

¦

-1.75

»¿

-1.72

¡

-1.72

±

-1.59

ught

-1.58

»

-1.58

²

-1.57

¢

-1.55

¬

-1.54

POSITIVE LOGITS

 whereby

1.71

 booth

1.56

ession

1.52

 operative

1.49

heets

1.46

walks

1.46

pace

1.43

 device

1.43

forge

1.42

IZE

1.42

Activations Density 0.046%

definitions or meanings of terms

No Comments

No Known Activations