INDEX

Explanations

terms related to favoritism and selection preferences

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

 respectively

-1.58

ogether

-1.36

vent

-1.36

ynes

-1.33

#####

-1.29

tract

-1.29

using

-1.24

--------------------------------------------------------------------------------------------------------------------------------

-1.23

ophyll

-1.23

etc

-1.23

POSITIVE LOGITS

ħ

2.30

¦

2.22

ŀ

2.18

©

2.13

ĥ½

2.11

ī

2.06

¹

2.06

ģ

2.01

Ŀ

2.00

¤

1.99

Activations Density 0.199%

terms related to favoritism and selection preferences

No Comments

No Known Activations