INDEX

Explanations

No Explanations Found

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

</

-1.89

ld

-1.64

}</

-1.54

Sprintf

-1.53

============

-1.51

===============

-1.48

}}{\

-1.39

 promotes

-1.39

 solves

-1.36

 depends

-1.36

POSITIVE LOGITS

¢

2.32

ĳ

2.30

ŀ

2.19

µ

2.12

ł

1.92

Ń

1.83

¦

1.82

ľ

1.81

½

1.79

ľĵ

1.77

Activations Density 0.000%

No Known Activations

This feature has no known activations.