INDEX

Explanations

Code, documentation

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_24.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.24.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

èĳ«

-0.32

ä¸Ģç«Ļ

-0.30

Ecc

-0.27

æ´ĭæ´ĭ

-0.27

 pÃ¡

-0.26

pin

-0.25

æĹ©æĻļ

-0.25

ecc

-0.25

èĲģ

-0.24

noinspection

-0.23

POSITIVE LOGITS

 radicals

0.25

çļĦçĶ·äºº

0.25

fel

0.25

 Kurdish

0.24

åĨ²çªģ

0.24

ure

0.23

èĩ´

0.23

èµ·é£ŀ

0.23

 Culture

0.23

ural

0.23

Activations Density 0.014%

Code, documentation

No Comments

No Known Activations