INDEX

Explanations

Names, organizations, misspellings

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_27.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.27.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

éĴº

-0.27

 donate

-0.25

ppl

-0.24

_kill

-0.24

WARDED

-0.24

Ø±Ø§Øª

-0.24

Ð¸ÑĤÐµÐ»ÑĮ

-0.24

Ð¸ÑĤÐµÐ»ÑĮÐ½Ð¾

-0.24

åĨħç½®

-0.24

wear

-0.23

POSITIVE LOGITS

åĮĹæĸ¹

0.25

@js

0.25

od

0.25

ï¼ĮãĢĬ

0.24

 Classified

0.24

 almost

0.24

 rusty

0.24

ä¹ĭåĬĽ

0.23

 Northern

0.23

 obst

0.23

Activations Density 0.003%

Names, organizations, misspellings

No Known Activations