INDEX

Explanations

phrases indicating the presence of conditions or traits in subjects

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

»¿

-2.04

isn

-1.93

forum

-1.85

ption

-1.85

Ī

-1.77

Ľ

-1.76

facts

-1.75

»

-1.75

ptions

-1.74

endas

-1.73

POSITIVE LOGITS

 evenly

2.31

 respectively

2.01

 randomly

1.95

 kindly

1.88

hers

1.85

 initially

1.85

 trypsin

1.81

hes

1.81

 agarose

1.78

 sequentially

1.74

Activations Density 0.231%

phrases indicating the presence of conditions or traits in subjects

No Comments

No Known Activations