INDEX

Explanations

instances of phrases involving the concept of "to" followed by various verbs or actions

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/2-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.2.hook_resid_post

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĩ

-3.36

¡

-3.32

¿½

-3.29

¼

-3.25

ĸ

-3.23

Ĺ

-3.21

µ

-3.21

±

-3.18

Ĳ

-3.18

Ĩ

-3.13

POSITIVE LOGITS

akes

1.67

osure

1.67

 Redistributions

1.55

ashi

1.52

portation

1.51

iled

1.49

 omit

1.48

 slice

1.45

geries

1.43

rous

1.40

Activations Density 0.584%

instances of phrases involving the concept of "to" followed by various verbs or actions

No Comments

No Known Activations