INDEX

Explanations

statements indicating the nature of comparisons or evaluations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-res-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.1.hook_resid_post

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

 vigil

-1.65

aring

-1.56

 liberty

-1.40

 dedicated

-1.39

 alive

-1.38

inde

-1.37

 courts

-1.36

 discovers

-1.34

den

-1.34

usal

-1.33

POSITIVE LOGITS

ORT

1.42

 pony

1.41

 pairing

1.38

aceae

1.35

 aside

1.34

 Ïī

1.31

 correlate

1.31

 vocals

1.31

åĬĽ

1.31

oS

1.31

Activations Density 0.222%

statements indicating the nature of comparisons or evaluations

No Comments

No Known Activations