INDEX

Explanations

Beginnings of titles

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_24.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.24.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 essence

-0.30

ç»©æķĪ

-0.28

çĤ¹è¯Ħ

-0.27

å¿ħè¦ģ

-0.27

çļĦæľ¬è´¨

-0.27

éº»è¾£

-0.27

 lockdown

-0.26

Ess

-0.26

æī§æĦı

-0.26

 essential

-0.26

POSITIVE LOGITS

atoi

0.26

 setId

0.26

åħ¬ç¤¾

0.26

 Seymour

0.25

 Except

0.25

 Alger

0.25

')(

0.25

 Scores

0.24

except

0.24

 ConfigureServices

0.23

Activations Density 0.163%

Beginnings of titles

No Known Activations