INDEX

Explanations

Codes and Identifiers

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ç§°

-0.28

RuleContext

-0.28

charging

-0.26

 concept

-0.25

.fi

-0.25

ä»»ä½ķå½¢å¼ı

-0.24

 narrower

-0.24

 yourselves

-0.24

DMETHOD

-0.24

äºĽä»Ģä¹Ī

-0.23

POSITIVE LOGITS

gov

0.28

elon

0.27

è¾Ĳ

0.26

_PROPERTY

0.26

æ±½

0.26

æĪĴæĮĩ

0.26

Inst

0.25

 Dund

0.24

_ph

0.24

Bulletin

0.24

Activations Density 0.039%

Codes and Identifiers

No Comments

No Known Activations