INDEX

Explanations

willing

np_max-act · gemini-2.0-flash

concepts related to success and achievement.

oai_token-act-pair · gpt-4o-mini Triggered by @xinyanhu8

The neuron consistently activates on occurrences of the word “willing,” i.e. expressions of readiness or consent.

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

 canon

-0.07

 효과

-0.07

intendo

-0.07

 Scotland

-0.07

ApiModelProperty

-0.07

 Coupe

-0.07

 měsíců

-0.07

AO

-0.07

 Century

-0.07

 захисту

-0.07

POSITIVE LOGITS

 willing

0.13

 willingness

0.10

 unwilling

0.08

 unable

0.07

lie

0.07

lin

0.07

 Allison

0.07

etine

0.07

 взаим

0.07

 Offering

0.06

Activations Density 0.006%

willing

concepts related to success and achievement.

The neuron consistently activates on occurrences of the word “willing,” i.e. expressions of readiness or consent.

No Comments

No Known Activations