INDEX

Explanations

Code, technical content

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_27.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.27.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

 RelayCommand

-0.27

eros

-0.27

æ¾¹

-0.26

ä¸Ģä»¶

-0.26

/article

-0.25

æĺ¯ä¸Ģä»¶

-0.25

Ĳľç´¢

-0.25

observation

-0.24

MES

-0.24

 KhÃ¡ch

-0.24

POSITIVE LOGITS

ä½ĵåĪ¶

0.34

æĢĿç»´æĸ¹å¼ı

0.32

æĢĿç»´

0.30

 environment

0.30

 lifestyle

0.30

coni

0.29

 society

0.28

è·¯çº¿

0.28

 personalities

0.28

æĹ¥åŃĲ

0.28

Activations Density 0.000%

Code, technical content

No Comments

No Known Activations