INDEX

Explanations

may

np_max-act · gemini-2.0-flash

This neuron detects permissive modal verbs (especially “may”) in legal or formal text.

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

Brown

-0.06

Dj

-0.06

 conte

-0.06

=c

-0.06

*v

-0.06

 اینجا

-0.06

 Wanna

-0.06

雑

-0.06

Friend

-0.06

 سعر

-0.05

POSITIVE LOGITS

.getLeft

0.08

 formulate

0.07

Serializable

0.07

Drawable

0.06

 extrem

0.06

	form

0.06

ismic

0.06

iddled

0.06

 gi�

0.06

 Türkçe

0.06

Activations Density 0.026%

may

This neuron detects permissive modal verbs (especially “may”) in legal or formal text.

No Comments

No Known Activations

may

This neuron detects permissive modal verbs (especially “may”) in legal or formal text.

No Comments

No Known Activations