INDEX

Explanations

Staying abreast or approximation

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_2.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.2.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

èħĬ

-0.25

 youre

-0.24

/*/

-0.24

à¤¦

-0.24

mart

-0.24

_puts

-0.23

èĪ¬

-0.23

LOUR

-0.23

æĸĳ

-0.23

zwy

-0.23

POSITIVE LOGITS

nda

0.27

lo

0.26

ntity

0.26

 stabilized

0.25

atoria

0.25

ç¨³

0.25

æ³ķå¾ĭè´£ä»»

0.24

osl

0.24

åģľæ»ŀ

0.24

hidden

0.24

Activations Density 37.571%

Staying abreast or approximation

No Comments

No Known Activations