INDEX

Explanations

Code and programming

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_23.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.23.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

ospace

-0.27

atile

-0.26

 occupying

-0.25

efd

-0.24

 Orient

-0.24

è¹Ĭ

-0.24

æĬµæĬ¼

-0.24

<Result

-0.23

 subdiv

-0.23

 speculation

-0.23

POSITIVE LOGITS

_rhs

0.29

ptron

0.28

Ã£ng

0.26

ä¸»åĭķ

0.25

prior

0.24

manship

0.24

ialis

0.24

åĨħå®¹ç®Ģä»ĭ

0.24

æĦŁ

0.24

æĦıç¾©

0.24

Activations Density 0.022%

Code and programming

No Comments

No Known Activations