INDEX

Explanations

code, programming

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_35.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.35.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 emits

-0.39

 emitted

-0.37

 carbon

-0.34

 emit

-0.33

emit

-0.32

 emitting

-0.32

åıĳå°Ħ

-0.31

 emission

-0.30

	emit

-0.30

ç¢³

-0.30

POSITIVE LOGITS

bis

0.35

bing

0.32

bia

0.30

bcc

0.29

bias

0.28

ber

0.28

ping

0.28

libraries

0.28

0.27

bor

0.27

Activations Density 0.011%

code, programming

No Known Activations