INDEX

Explanations

.add

np_max-act · gemini-2.0-flash

The neuron selectively activates on pptx “.add_…” method calls (e.g. add_slide, add_textbox, add_picture), i.e. tokens marking the addition of slides or shapes in Python‐pptx code.

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

 visas

-0.07

_pod

-0.06

 декіль

-0.06

 redesign

-0.06

 метал

-0.06

 Tire

-0.06

sip

-0.06

 المر

-0.06

Soup

-0.06

ATAL

-0.06

POSITIVE LOGITS

 추가

0.07

において

0.07

 Keyword

0.06

 witnessed

0.06

 사람

0.06

seb

0.06

 asıl

0.06

ac

0.06

.setDate

0.06

 manus

0.06

Activations Density 0.014%

.add

The neuron selectively activates on pptx “.add_…” method calls (e.g. add_slide, add_textbox, add_picture), i.e. tokens marking the addition of slides or shapes in Python‐pptx code.

No Comments

No Known Activations