INDEX

Explanations

Law of Attraction

np_max-act · gemini-2.0-flash

The neuron detects the character sequence “tract” (and its variants, e.g. in “attraction,” “atração,” “vision-board”) whenever words containing that root appear.

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

"))

-0.07

"))

-0.07

}}>

-0.06

的话

-0.06

ican

-0.06

 IsNot

-0.06

onenumber

-0.06

KY

-0.06

_PATTERN

-0.06

icas

-0.06

POSITIVE LOGITS

thumbs

0.07

Events

0.06

Nib

0.06

Comment

0.06

.Val

0.06

 woodworking

0.06

 Concern

0.06

 Draft

0.06

Atoms

0.06

 Chow

0.06

Activations Density 0.035%

Law of Attraction

The neuron detects the character sequence “tract” (and its variants, e.g. in “attraction,” “atração,” “vision-board”) whenever words containing that root appear.

No Comments

No Known Activations