INDEX

Explanations

clear

np_max-act · gemini-2.0-flash

This neuron activates on words flagging that the user’s question or statement is unclear or vague (e.g. “clear” in “not clear how…”).

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Services

-0.07

 Selection

-0.06

(""+

-0.06

Bah

-0.06

 Sorted

-0.06

equals

-0.06

提

-0.06

 thư

-0.06

和

-0.06

 ประเทศ

-0.06

POSITIVE LOGITS

 تصم

0.07

 clearer

0.07

 done

0.07

 української

0.07

时

0.06

ункт

0.06

 góc

0.06

今年

0.06

'):
↵

0.06

क

0.06

Activations Density 0.013%

clear

This neuron activates on words flagging that the user’s question or statement is unclear or vague (e.g. “clear” in “not clear how…”).

No Comments

No Known Activations