INDEX

Explanations

bless

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_31.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.31.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

è¸ıä¸Ĭ

-0.26

ÙĬØ¨

-0.25

åıªçŁ¥éģĵ

-0.25

åĽ¡

-0.25

 Hawks

-0.24

-interface

-0.24

æĶ¾å¤§

-0.24

åºŀ

-0.24

Care

-0.24

theast

-0.23

POSITIVE LOGITS

 given

0.34

ç»ĻäºĪ

0.29

given

0.29

 above

0.28

èµĲ

0.28

çµ¦äºĪ

0.28

ç»ĻäºĪäºĨ

0.27

åħį

0.26

ä¾Ľæ°´

0.26

 provided

0.26

Activations Density 0.132%

bless

No Comments

No Known Activations