INDEX

Explanations

seeking medical attention

np_max-act · gemini-2.0-flash

This neuron activates on tokens within advisory or instructional statements—especially health-care recommendations and directives (e.g., “you,” “should,” “contact,” “can,” etc.).

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

 unions

-0.07

 woll

-0.07

ebp

-0.07

 إليه

-0.07

 yapılması

-0.06

ريقة

-0.06

fg

-0.06

ijn

-0.06

双

-0.06

	connection

-0.06

POSITIVE LOGITS

0.07

.Gradient

0.07

 mosquito

0.06

 releasing

0.06

 Hose

0.06

 läng

0.06

uyển

0.06

 Dân

0.06

 Previously

0.06

 Humb

0.06

Activations Density 0.020%

seeking medical attention

This neuron activates on tokens within advisory or instructional statements—especially health-care recommendations and directives (e.g., “you,” “should,” “contact,” “can,” etc.).

No Comments

No Known Activations