INDEX

Explanations

studied

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

çķĮçļĦ

-0.27

å¤Ļ

-0.27

 tats

-0.27

éĺħ

-0.27

odos

-0.25

lick

-0.25

æĻ®æŁ¥

-0.25

æĻ¾

-0.25

çĥĺå¹²

-0.25

 Ones

-0.25

POSITIVE LOGITS

python

0.28

Python

0.28

Å¼u

0.27

ä¸Ńæĸĩ

0.26

.python

0.25

å¿ĥçĲĨåŃ¦

0.25

åįļå£«åŃ¦ä½į

0.24

 Python

0.24

 away

0.24

 python

0.24

Activations Density 0.005%

studied

No Comments

No Known Activations