INDEX

Explanations

-small

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_3/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.3.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 водой

-0.07

 güneş

-0.06

دم

-0.06

 ignorance

-0.06

 nhớ

-0.06

 yapılan

-0.06

φ

-0.06

 después

-0.06

Lal

-0.06

POSITIVE LOGITS

 vegetarian

0.11

 Ebola

0.11

-small

0.09

 MERCHANTABILITY

0.08

 cellpadding

0.08

 masturbation

0.07

 Cele

0.06

 теб

0.06

 ноя

0.06

 onActivityResult

0.06

Activations Density 0.004%

-small

No Comments

No Known Activations