INDEX

Explanations

maintain/keep

np_max-act · gemini-2.0-flash

This neuron activates on words that express preserving or sustaining something—e.g. “maintain,” “keep,” “remain.”

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.08

-instance

-0.07

次

-0.07

เท

-0.07

	table

-0.07

/install

-0.07

 begin

-0.06

rama

-0.06

 pointed

-0.06

POSITIVE LOGITS

 enrich

0.06

 корп

0.06

属于

0.06

 smoothly

0.06

ุบาล

0.06

 sustained

0.06

_;↵

0.06

لة

0.06

_ASM

0.06

@register

0.06

Activations Density 0.120%

maintain/keep

This neuron activates on words that express preserving or sustaining something—e.g. “maintain,” “keep,” “remain.”

No Comments

No Known Activations