INDEX

Explanations

discrimination

np_max-act · gemini-2.0-flash

This neuron fires on words that specify discriminatory reasons—i.e. the “because of ‹attribute›” constructions naming who someone is (e.g. who they love, their religion, sexual orientation, nationality) as grounds for unjust treatment.

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

stru

-0.07

푸

-0.07

 Αθή

-0.06

 Crom

-0.06

 کشور

-0.06

JNIEnv

-0.06

returned

-0.06

учас

-0.06

_wind

-0.06

 ци

-0.06

POSITIVE LOGITS

 Billboard

0.06

 cins

0.06

NL

0.06

CFG

0.06

 Disposable

0.06

 правиль

0.06

 Executive

0.06

 bcrypt

0.06

SVG

0.06

 cooker

0.06

Activations Density 0.013%

discrimination

This neuron fires on words that specify discriminatory reasons—i.e. the “because of ‹attribute›” constructions naming who someone is (e.g. who they love, their religion, sexual orientation, nationality) as grounds for unjust treatment.

No Comments

No Known Activations

discrimination

This neuron fires on words that specify discriminatory reasons—i.e. the “because of ‹attribute›” constructions naming who someone is (e.g. who they love, their religion, sexual orientation, nationality) as grounds for unjust treatment.

No Comments

No Known Activations