INDEX

Explanations

terms and phrases related to punishment and consequences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-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.1.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

»¿

-2.94

Ŀ

-2.62

¿½

-2.53

Ī

-2.42

ħ

-2.41

ĥ½

-2.40

ĻĤ

-2.38

ª

-2.32

ı

-2.31

↵

-2.19

POSITIVE LOGITS

 regime

2.00

 penalties

1.88

 punish

1.86

 violations

1.85

 fines

1.78

 regimes

1.76

 amount

1.73

 ordinance

1.68

 ordin

1.64

 sanctions

1.62

Activations Density 0.432%

terms and phrases related to punishment and consequences

No Comments

No Known Activations