INDEX

Explanations

Code and technical discussions

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

Negative Logits

ä¼ĳ

-0.29

æĦıåĳ³çĿĢ

-0.28

hone

-0.27

 prof

-0.26

 unions

-0.25

åĪ©çĽĬ

-0.25

 priv

-0.25

 franca

-0.24

æĴŃ

-0.24

å¤§å¤§

-0.24

POSITIVE LOGITS

 Came

0.28

IGNAL

0.26

 Trial

0.24

ctest

0.24

 evapor

0.24

 smashed

0.24

ibil

0.23

aggi

0.23

ailed

0.23

 berlin

0.23

Activations Density 0.000%

Code and technical discussions

No Known Activations