INDEX

Explanations

spec

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_22.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.22.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

asc

-0.26

oss

-0.25

erm

-0.25

ene

-0.24

anh

-0.24

popular

-0.24

azon

-0.24

 popular

-0.23

ahoma

-0.23

åıĳçĶŁåľ¨

-0.23

POSITIVE LOGITS

imen

0.29

.layouts

0.27

ãĤĪ

0.24

 Jeans

0.24

æĸ

0.24

ÑĨÐ¸Ñı

0.24

thesis

0.24

ifier

0.23

ulle

0.23

ä»

0.23

Activations Density 0.179%

spec

No Known Activations