INDEX

Explanations

harm, injury, death

np_max-act · gemini-2.0-flash

This neuron responds to words that denote serious harm or adverse outcomes—e.g., injuries, morbidity, hospitalization, catastrophic events.

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

ับม

-0.07

 Distributed

-0.07

Large

-0.06

ion

-0.06

onic

-0.06

击

-0.06

ewart

-0.06

 tactical

-0.06

 deem

-0.06

 asteroid

-0.06

POSITIVE LOGITS

?>">↵

0.07

 dům

0.06

 ullam

0.06

_mul

0.06

 Armor

0.06

落

0.06

during

0.06

-awesome

0.06

*))

0.06

 BOARD

0.06

Activations Density 0.132%

harm, injury, death

This neuron responds to words that denote serious harm or adverse outcomes—e.g., injuries, morbidity, hospitalization, catastrophic events.

No Comments

No Known Activations