INDEX

Explanations

magazine

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

 Belmont

-0.27

 WebClient

-0.27

äº¤èŀį

-0.25

strom

-0.25

sono

-0.25

 Retreat

-0.24

bestos

-0.24

 ProgressBar

-0.24

 Corbyn

-0.24

æŃ£ä¹ī

-0.24

POSITIVE LOGITS

 publication

0.32

 subscription

0.28

 publishers

0.28

åĩºçīĪ

0.27

 presentation

0.27

 published

0.27

 ç¼

0.26

 publisher

0.26

 creation

0.26

 cover

0.25

Activations Density 0.138%

magazine

No Comments

No Known Activations