INDEX

Explanations

phrases indicating value or worth

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

oder

-1.74

ores

-1.59

Ł

-1.59

ause

-1.51

Donnell

-1.51

ech

-1.49

Ĩ

-1.47

STITUTE

-1.44

 Schools

-1.43

respectively

-1.43

POSITIVE LOGITS

iness

2.12

lessly

2.06

iest

1.89

ies

1.89

ful

1.85

lessness

1.67

ily

1.66

fully

1.58

'[

1.57

chitz

1.56

Activations Density 0.268%

phrases indicating value or worth

No Comments

No Known Activations