INDEX

Explanations

ethics

np_max-act · gemini-2.0-flash

The neuron flags normative or ethical‐judgment words (like “illegal,” “unethical,” “responsible,” “safe,” etc.) in the 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

کز

-0.08

preg

-0.07

 kurs

-0.07

 نس

-0.07

Calls

-0.06

 Ев

-0.06

 розрах

-0.06

ارش

-0.06

 марш

-0.06

 Rebels

-0.06

POSITIVE LOGITS

ited

0.08

essaging

0.07

TOP

0.06

"*"

0.06

 Once

0.06

strings

0.06

/entity

0.06

$_[

0.06

bk

0.06

 showed

0.06

Activations Density 0.013%

ethics

The neuron flags normative or ethical‐judgment words (like “illegal,” “unethical,” “responsible,” “safe,” etc.) in the text.

No Comments

No Known Activations