INDEX

Explanations

references to array and tensor data structures

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

^âĪĴ

-1.56

^--

-1.53

MSO

-1.51

 Citizens

-1.50

?"

-1.47

>/

-1.44

>=

-1.44

...?"

-1.43

 Workers

-1.41

>"

-1.39

POSITIVE LOGITS

borne

1.85

ÅĽci

1.83

eries

1.82

bourg

1.81

etable

1.74

wordpress

1.67

leen

1.63

bag

1.61

enstein

1.60

nai

1.59

Activations Density 0.014%

references to array and tensor data structures

No Comments

No Known Activations