INDEX

Explanations

technical text

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_27.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.27.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anga

-0.29

ÑīÐµÐ½

-0.28

 Äĳá»ķ

-0.27

éĩį

-0.26

aman

-0.26

uto

-0.26

æŃ£ä¹ī

-0.25

Rad

-0.25

odia

-0.25

åį«è§Ĩ

-0.24

POSITIVE LOGITS

 barracks

0.30

.air

0.29

emachine

0.28

ZX

0.26

åĨįä¹Łæ²¡æľī

0.26

fabric

0.25

HandlerContext

0.25

dle

0.24

ertext

0.24

ä¸ĩåĲį

0.24

Activations Density 0.000%

technical text

No Comments

No Known Activations