INDEX

Explanations

repeated occurrences of the same element or pattern

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-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_mlp_out

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

·¸

-1.66

Ļ

-1.57

ķ

-1.55

actors

-1.52

 wonders

-1.40

 mention

-1.38

eyed

-1.38

 plans

-1.37

 predictions

-1.36

¢

-1.36

POSITIVE LOGITS

+^

1.62

↵ ↵

1.57

%.

1.56

eds

1.53

zyk

1.52

igraph

1.51

/~

1.47

ary

1.46

phab

1.45

ki

1.45

Activations Density 2.888%

repeated occurrences of the same element or pattern

No Comments

No Known Activations