INDEX

Explanations

phrases indicating durations or time periods

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__att-sm_processed/0-att-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.0.hook_attn_out

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

imony

-1.72

ĥ

-1.52

Ļª

-1.50

iest

-1.47

ily

-1.45

sing

-1.40

 accusations

-1.35

vou

-1.34

ifications

-1.33

liest

-1.31

POSITIVE LOGITS

Rptr

1.58

à«ĩ

1.57

estock

1.56

================================================================

1.54

↵ ↵

1.52

doms

1.48

 Rptr

1.46

BSD

1.45

eon

1.43

==============================

1.40

Activations Density 0.441%

phrases indicating durations or time periods

No Comments

No Known Activations