INDEX

Explanations

terms relating to micro and macro concepts in various contexts

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

 msgid

-1.73

 willingness

-1.57

rapeut

-1.55

juvant

-1.52

arer

-1.51

ocarcinoma

-1.47

andidates

-1.46

herty

-1.45

REAM

-1.43

UTF

-1.42

POSITIVE LOGITS

cles

1.76

ħ

1.52

 screens

1.51

Ĵ

1.49

 artifacts

1.48

chip

1.45

Ħ

1.45

(\~

1.44

 artifact

1.41

 chips

1.40

Activations Density 0.061%

terms relating to micro and macro concepts in various contexts

No Comments

No Known Activations