INDEX

Explanations

adopt

np_max-act · gemini-2.0-flash

The neuron consistently flags instances of “adopt” or “adoption” (and their inflections), i.e. words conveying the concept of adopting.

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

 humour

-0.07

 area

-0.07

 humor

-0.07

 experiencia

-0.07

 Bitcoins

-0.07

 circum

-0.07

Fi

-0.07

 conduit

-0.07

 Freund

-0.06

Sim

-0.06

POSITIVE LOGITS

 adopt

0.14

 adopting

0.13

 adopted

0.12

adopt

0.12

 Adopt

0.12

 adoption

0.12

 adopts

0.11

 Adoption

0.09

採

0.08

 infect

0.07

Activations Density 0.010%

adopt

The neuron consistently flags instances of “adopt” or “adoption” (and their inflections), i.e. words conveying the concept of adopting.

No Comments

No Known Activations

adopt

The neuron consistently flags instances of “adopt” or “adoption” (and their inflections), i.e. words conveying the concept of adopting.

No Comments

No Known Activations