INDEX

Explanations

references to news or news-related content

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

 myself

-2.23

§

-1.75

 mentally

-1.71

 himself

-1.70

 chemo

-1.61

 safely

-1.58

 healthy

-1.58

 overcome

-1.55

 resolved

-1.54

Ĳ

-1.53

POSITIVE LOGITS

letters

2.11

letter

2.10

wick

2.02

Channel

1.94

agent

1.91

CoV

1.90

room

1.84

week

1.84

flash

1.82

day

1.78

Activations Density 0.087%

references to news or news-related content

No Comments

No Known Activations