INDEX

Explanations

changes over time

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

IFrame

Link

Negative Logits

åıĹçĽĬ

-0.29

æĶ¹è¿Ľ

-0.28

iator

-0.27

éĵº

-0.27

 putchar

-0.27

æıŃç¤º

-0.26

asil

-0.26

æī¿åĬŀ

-0.26

äºĮç»´

-0.26

æĸ°äº¬

-0.26

POSITIVE LOGITS

 generation

0.28

 generations

0.28

istent

0.27

engan

0.27

:request

0.26

 decades

0.25

 Hairst

0.25

 Same

0.25

Same

0.25

 Generation

0.25

Activations Density 0.002%

changes over time

No Known Activations