INDEX

Explanations

you

np_max-act · gemini-2.0-flash

The neuron selectively lights up on second-person pronouns and related question-asking words (you, me, is, would, ask, etc.), i.e. when the model addresses or questions “you.”

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

 recurse

-0.07

Chuck

-0.06

入れ

-0.06

пон

-0.06

ію

-0.06

Ан

-0.06

ViewState

-0.06

är

-0.06

Hoa

-0.06

POSITIVE LOGITS

 پرو

0.06

 infiltr

0.06

_dest

0.06

 insider

0.06

.dataSource

0.06

i�

0.06

 generations

0.06

 progn

0.06

NEC

0.06

 исч

0.06

Activations Density 0.056%

you

The neuron selectively lights up on second-person pronouns and related question-asking words (you, me, is, would, ask, etc.), i.e. when the model addresses or questions “you.”

No Comments

No Known Activations