INDEX

Explanations

elements related to programming or technical references

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

ester

-1.35

imize

-1.35

CCT

-1.34

 Motors

-1.29

->

-1.27

ATM

-1.26

 homogen

-1.26

 normal

-1.25

 Tickets

-1.24

gra

-1.22

POSITIVE LOGITS

ãĤĥ

1.66

odies

1.48

uum

1.47

mons

1.42

umerable

1.41

à¯į

1.37

lessly

1.36

uities

1.35

ÑĥÐº

1.35

ifer

1.32

Activations Density 4.453%

elements related to programming or technical references

No Comments

No Known Activations