INDEX

Explanations

terms related to exploration and discovery

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

Negative Logits

acs

-1.68

oddsidemargin

-1.59

imes

-1.56

hers

-1.55

aho

-1.53

helm

-1.47

ensen

-1.45

arman

-1.44

oxide

-1.42

eless

-1.42

POSITIVE LOGITS

 depths

1.74

tain

1.61

unreadable

1.60

 feasibility

1.57

 tunnels

1.52

TeV

1.50

 gaps

1.43

 trap

1.41

how

1.39

 possibilities

1.38

Activations Density 0.439%

terms related to exploration and discovery

No Known Activations