INDEX

Explanations

Identifying topic or negation

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

éĥ½ä¸įæĺ¯

-0.32

è¿Ļä¸įæĺ¯

-0.31

ä¸įæĺ¯

-0.31

 neither

-0.30

ç¨įå¾®

-0.29

æ²¡æľīä»»ä½ķ

-0.29

è¿ĺä¸įæĺ¯

-0.29

ç¨įç¨į

-0.27

ä¸įä»£è¡¨

-0.27

 nowhere

-0.27

POSITIVE LOGITS

 kicks

0.28

 kick

0.27

åıĸ

0.25

{{{

0.24

kick

0.24

bras

0.23

caff

0.23

åħ³åĪĩ

0.23

 plays

0.23

itra

0.23

Activations Density 0.007%

Identifying topic or negation

No Known Activations