INDEX

Explanations

references to various types of buildings and locations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĥ½

-3.12

Ķ

-2.55

±

-2.52

¾

-2.52

¦

-2.46

Ŀ

-2.45

¤

-2.40

Ĳ

-2.36

º

-2.36

«

-2.36

POSITIVE LOGITS

owner

2.08

yard

2.07

 arrangement

1.94

fires

1.90

 arrangements

1.83

 system

1.81

 owner

1.78

 renovation

1.75

wright

1.75

holder

1.73

Activations Density 1.650%

references to various types of buildings and locations

No Comments

No Known Activations