INDEX

Explanations

suspension

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_30.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.30.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

igram

-0.27

rett

-0.26

éļĪ

-0.25

 trimming

-0.24

UTH

-0.24

appointed

-0.24

×¨×ļ

-0.24

ÑĤÐµÐº

-0.23

Mer

-0.23

åĩłå¼ł

-0.23

POSITIVE LOGITS

æĤ¬

1.15

 suspension

1.12

 susp

1.03

 Suspension

0.99

 suspend

0.90

æĤ¬æµ®

0.90

Susp

0.89

 suspended

0.88

æĩ

0.83

 Susp

0.82

Activations Density 1.069%

suspension

No Comments

No Known Activations