INDEX

Explanations

will

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 guidance

-0.30

æķĻ

-0.28

 advice

-0.27

å¯¹æĪĳ

-0.27

æĮĩå¯¼

-0.26

-dd

-0.25

ksz

-0.25

abbo

-0.24

ç«ĭæĸ¹

-0.24

 experience

-0.24

POSITIVE LOGITS

 Discussions

0.26

èħ¾

0.26

underscore

0.25

undra

0.25

å§Ĭå¦¹

0.25

è®®

0.25

ç®Ń

0.24

ä¸ĵéĹ¨

0.24

å¯¸

0.24

che

0.24

Activations Density 0.001%

will

No Comments

No Known Activations