INDEX

Explanations

function calls related to manipulating a stack or collection in programming

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

NHS

-1.69

 vivo

-1.46

 slightest

-1.43

 apparent

-1.42

 cross

-1.40

 adopt

-1.32

 known

-1.31

 contents

-1.28

 observable

-1.28

 observed

-1.27

POSITIVE LOGITS

manship

2.02

holder

1.99

oslav

1.85

starter

1.79

 legend

1.76

Ã±os

1.71

mania

1.68

ankind

1.67

Ã±o

1.66

ī

1.64

Activations Density 0.007%

function calls related to manipulating a stack or collection in programming

No Comments

No Known Activations