INDEX

Explanations

punctuation marks

np_max-act · gemini-2.0-flash

The neuron primarily activates on common function words and question/instruction markers (e.g. prepositions like “on,” “in,” connectors like “and,” modal/infinitive forms like “can,” “to,” and interrogatives like “what,” “how”).

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mpeg

-0.08

OfBirth

-0.06

mph

-0.06

onacci

-0.06

 antibiotic

-0.06

 crash

-0.06

myfile

-0.06

	stack

-0.06

_LS

-0.06

_cod

-0.06

POSITIVE LOGITS

setCurrent

0.07

 هستند

0.07

;o

0.07

WON

0.07

 НА

0.07

whether

0.07

register

0.06

(""))

0.06

.setItem

0.06

点击

0.06

Activations Density 0.097%

punctuation marks

The neuron primarily activates on common function words and question/instruction markers (e.g. prepositions like “on,” “in,” connectors like “and,” modal/infinitive forms like “can,” “to,” and interrogatives like “what,” “how”).

No Comments

No Known Activations