INDEX

Explanations

must

np_max-act · gemini-2.0-flash

This neuron detects strong obligation or requirement language, i.e. modal cues like “must,” “have to,” “mandatory,” or “follow.”

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

flo

-0.07

elerinin

-0.06

 Richards

-0.06

DA

-0.06

ぼ

-0.06

 differed

-0.06

 laat

-0.06

fc

-0.06

 Wheeler

-0.06

Wide

-0.06

POSITIVE LOGITS

(prog

0.07

URLConnection

0.07

ATCH

0.07

Luo

0.06

Impact

0.06

agnetic

0.06

IRMWARE

0.06

 років

0.06

іон

0.06

 unwind

0.06

Activations Density 0.026%

must

This neuron detects strong obligation or requirement language, i.e. modal cues like “must,” “have to,” “mandatory,” or “follow.”

No Comments

No Known Activations