INDEX

Explanations

health

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

 Power

-0.27

 Jung

-0.26

Power

-0.25

 Analytics

-0.25

Occ

-0.25

 Kang

-0.24

Top

-0.24

íĨµ

-0.24

ãĤĪãģĨ

-0.24

ç»Ł

-0.23

POSITIVE LOGITS

æĤ²åĵĢ

0.29

esian

0.26

 stew

0.25

é©¬è·¯

0.25

å¦¹

0.25

uples

0.24

åĦĭ

0.24

æĤ²

0.24

 thinking

0.24

çĩ§

0.24

Activations Density 0.008%

health

No Comments

No Known Activations