INDEX

Explanations

research papers and linguistics

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_31.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.31.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

iska

-0.27

abus

-0.25

 subj

-0.25

æºŁ

-0.24

alloca

-0.24

 sess

-0.24

odes

-0.23

 mastered

-0.23

ä¸´

-0.23

amma

-0.23

POSITIVE LOGITS

éĤ¸

0.28

ä¹ĭç§°

0.26

ball

0.26

bal

0.26

fan

0.25

åĽ´ç»ķ

0.25

ulin

0.25

è¯´å®ŀ

0.25

Com

0.24

æ²¸

0.24

Activations Density 0.919%

research papers and linguistics

No Known Activations