INDEX

Explanations

prime

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_30.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.30.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 focused

-0.28

minute

-0.27

 more

-0.26

more

-0.26

’en

-0.25

ns

-0.24

dialogs

-0.24

åħĪè¿Ľ

-0.24

legt

-0.24

ä¸ĢåĪ»

-0.24

POSITIVE LOGITS

-time

0.38

uem

0.32

-Time

0.31

owering

0.30

dating

0.30

 time

0.29

ean

0.29

çļĦæĹ¶éĹ´

0.28

time

0.27

eating

0.27

Activations Density 0.007%

prime

No Comments

No Known Activations