INDEX

Explanations

people

np_max-act · gemini-2.0-flash

This neuron activates on tokens indicating speaker roles or professional titles (e.g. “Industrial Designer,” “User Interface,” “Project Manager,” 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

Roc

-0.07

 Instagram

-0.06

 started

-0.06

 tragic

-0.06

Ob

-0.06

started

-0.06

 Sequ

-0.06

 Mature

-0.06

 Borrow

-0.06

POSITIVE LOGITS

 soudu

0.08

�

0.07

 uży

0.07

 quarterly

0.07

姓名

0.07

 зм

0.07

 zamanda

0.07

ارات

0.07

_COMPLEX

0.07

重大

0.06

Activations Density 0.016%

people

This neuron activates on tokens indicating speaker roles or professional titles (e.g. “Industrial Designer,” “User Interface,” “Project Manager,” etc.).

No Comments

No Known Activations