INDEX

Explanations

updates

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_25.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.25.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

obia

-0.31

crement

-0.27

 batting

-0.27

éĺĸ

-0.27

/platform

-0.27

éĿ¢ç§¯çº¦

-0.26

 ((((

-0.26

VERRIDE

-0.26

æĳĨåľ¨

-0.26

æľĶ

-0.25

POSITIVE LOGITS

 updates

0.29

æĽ´æĸ°

0.29

æĽ´å¤ļ

0.28

 redo

0.28

éĩįæĸ°

0.27

re

0.27

 restructuring

0.25

çĸı

0.25

 æĽ´æĸ°

0.25

 translations

0.25

Activations Density 0.907%

updates

No Known Activations