INDEX

Explanations

Code and medical texts

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

ÃŃch

-0.29

chin

-0.28

ogen

-0.27

operand

-0.27

 prosecute

-0.26

 serviced

-0.26

 goodness

-0.26

 execute

-0.25

èº«

-0.25

 Lifetime

-0.25

POSITIVE LOGITS

 symptoms

0.35

è¯¯åĮº

0.31

çļĦçĹĩçĬ¶

0.31

 Symptoms

0.30

çĹĩçĬ¶

0.29

 symptom

0.27

ä»Ģä¹ĪæĹ¶åĢĻ

0.27

æ³¨æĦı

0.26

 scenarios

0.26

agnetic

0.25

Activations Density 0.001%

Code and medical texts

No Comments

No Known Activations