INDEX

Explanations

the word "to" in various contexts

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

 retrospect

-1.34

 psychiat

-1.33

orms

-1.32

my

-1.29

GIS

-1.26

Äį

-1.23

 logging

-1.22

romycin

-1.20

 normalized

-1.20

 depression

-1.20

POSITIVE LOGITS

»¿

2.03

ĻĤ

1.92

ĨĴ

1.70

ĩ

1.67

inel

1.66

½

1.65

 therefrom

1.62

Ĩ

1.61

ŀ

1.50

©

1.49

Activations Density 1.625%

the word "to" in various contexts

No Comments

No Known Activations