INDEX

Explanations

instances of positive feedback or praise

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

ľ

-2.10

¾

-2.04

Ŀ

-2.04

³

-1.99

º

-1.94

į

-1.92

ı

-1.84

¡

-1.82

µ

-1.79

ª

-1.78

POSITIVE LOGITS

 Code

1.59

 species

1.56

 Species

1.50

 others

1.48

 sexes

1.43

","

1.38

itone

1.37

 Parenthood

1.36

 please

1.36

 Others

1.36

Activations Density 0.033%

instances of positive feedback or praise

No Comments

No Known Activations