INDEX

Explanations

occurrences of the word "FROM."

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

 hydroxyl

-1.49

 traced

-1.46

atin

-1.42

 brow

-1.41

'?

-1.38

'?"

-1.36

âĢī

-1.35

'$

-1.32

res

-1.31

hem

-1.31

POSITIVE LOGITS

heed

1.92

anks

1.65

 brakes

1.65

thouse

1.59

 draft

1.55

aine

1.54

ESULT

1.52

aylor

1.52

 Rptr

1.51

 bushes

1.51

Activations Density 0.138%

occurrences of the word "FROM."

No Comments

No Known Activations