INDEX

Explanations

to

np_max-act · gemini-2.0-flash

This neuron activates on the word “able,” i.e. constructions expressing that something is capable of or able to do something.

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

 editions

-0.06

 прос

-0.06

 bgColor

-0.06

_red

-0.06

dor

-0.06

 hips

-0.06

ávky

-0.06

 ylim

-0.06

 communities

-0.06

issue

-0.05

POSITIVE LOGITS

 tabela

0.08

 První

0.07

είται

0.07

 enact

0.07

 >↵

0.07

_piece

0.07

 sera

0.07

 guit

0.07

 :↵↵

0.07

>',↵

0.06

Activations Density 0.030%

to

This neuron activates on the word “able,” i.e. constructions expressing that something is capable of or able to do something.

No Comments

No Known Activations