INDEX

Explanations

Medical/scientific publications

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_7.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.7.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 impe

-0.28

æ³®

-0.28

 clap

-0.27

ordova

-0.27

oji

-0.26

ä¸ļ

-0.26

èµĽåŃ£

-0.25

orns

-0.25

IMIZE

-0.24

occus

-0.24

POSITIVE LOGITS

 Rather

0.28

 mall

0.27

 somewhere

0.26

 instead

0.26

æ¼«

0.25

instead

0.25

å¿½æĤł

0.25

WER

0.25

 Mall

0.24

Activations Density 0.001%

Medical/scientific publications

No Comments

No Known Activations