INDEX

Explanations

reconstruction

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ä»Ģä¹ĪåİŁåĽł

-0.27

åħ»

-0.25

 TestCase

-0.24

vant

-0.24

nement

-0.24

 unlucky

-0.24

åı£

-0.23

ned

-0.23

edList

-0.23

 behalf

-0.23

POSITIVE LOGITS

Cpp

0.27

lette

0.27

åĪĨæµģ

0.26

èĬĤæ°´

0.26

orer

0.26

 Erot

0.24

psz

0.24

å¾Ģåīį

0.24

-eff

0.23

aget

0.23

Activations Density 0.007%

reconstruction

No Comments

No Known Activations