INDEX

Explanations

programming-related terms and keywords

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

lain

-1.46

?**

-1.41

 completeness

-1.41

aland

-1.36

 Guarant

-1.36

eligible

-1.35

.**

-1.34

]"

-1.34

Son

-1.30

 contradiction

-1.28

POSITIVE LOGITS

ĻĤ

2.58

£

2.57

§

2.47

Ľ

2.46

Īĺ

2.40

ģ

2.38

ĳ

2.34

ĺ

2.31

ĭ

2.31

±

2.30

Activations Density 0.682%

programming-related terms and keywords

No Comments

No Known Activations