INDEX

Explanations

writing excerpts

np_max-act · gemini-2.0-flash

This neuron detects mentions of document-structure instructions that specify section names together with quantitative details (e.g., “Introduction of 3 pages”).

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

 miscar

-0.07

 chants

-0.07

flation

-0.07

 shock

-0.06

Tour

-0.06

Hor

-0.06

 vector

-0.06

 born

-0.06

 episode

-0.06

 ποι

-0.06

POSITIVE LOGITS

 سی

0.07

pecia

0.07

์ฟ

0.06

 Stim

0.06

 tướng

0.06

FindBy

0.06

 Степ

0.06

_cond

0.06

 Beirut

0.06

Activations Density 0.058%

writing excerpts

This neuron detects mentions of document-structure instructions that specify section names together with quantitative details (e.g., “Introduction of 3 pages”).

No Comments

No Known Activations