INDEX

Explanations

Expressing and communicating ideas

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_20.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.20.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ulously

-0.32

çĺ¾

-0.30

å®ŀçī©

-0.28

inea

-0.28

èľĺèĽĽ

-0.26

bane

-0.25

 flawed

-0.25

LL

-0.24

ãģįãģ¡ãĤĵ

-0.24

ells

-0.23

POSITIVE LOGITS

 hookup

0.30

 slightest

0.29

 presence

0.27

presence

0.25

egie

0.25

 whatsoever

0.25

 slur

0.25

 wink

0.24

ç¾¹

0.24

oster

0.24

Activations Density 22.851%

Expressing and communicating ideas

No Known Activations