INDEX

Explanations

Unusual or fragmented text

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_19.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.19.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Yourself

-0.27

WebService

-0.26

letics

-0.25

ksam

-0.25

ä¼Ĭå§ĭ

-0.25

èĥ½æĬĬ

-0.25

ariat

-0.25

Ã©sz

-0.24

.hasNext

-0.24

nad

-0.24

POSITIVE LOGITS

 represents

0.37

 intended

0.35

 meant

0.34

 intent

0.34

 represent

0.33

gil

0.33

ä»£è¡¨

0.32

intent

0.32

ä»£è¡¨çĿĢ

0.32

 intentions

0.30

Activations Density 0.223%

Unusual or fragmented text

No Comments

No Known Activations