INDEX

Explanations

backstory

np_max-act · gemini-2.0-flash

The neuron flags discussion of a character’s backstory and narrative setup (e.g. “backstory,” “story,” “events,” “past,” “experiences,” “motivations”).

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

Negative Logits

ewart

-0.07

 pořád

-0.06

 Тут

-0.06

 Bakanı

-0.06

ρχ

-0.06

!:

-0.06

icolor

-0.06

 ///↵

-0.06

!!!

-0.06

 ışık

-0.06

POSITIVE LOGITS

 backstory

0.08

研究

0.07

 sine

0.07

lete

0.07

ItemAt

0.07

=form

0.07

elm

0.07

 origin

0.07

.bs

0.06

_verified

0.06

Activations Density 0.021%

backstory

The neuron flags discussion of a character’s backstory and narrative setup (e.g. “backstory,” “story,” “events,” “past,” “experiences,” “motivations”).

No Known Activations

backstory

The neuron flags discussion of a character’s backstory and narrative setup (e.g. “backstory,” “story,” “events,” “past,” “experiences,” “motivations”).

No Known Activations