INDEX

Explanations

or

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_31.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.31.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

.Cast

-0.25

 over

-0.25

æķĻèĤ²åŁ¹è®Ń

-0.24

å¸ĤæķĻèĤ²

-0.24

uars

-0.24

æķĻ

-0.24

iron

-0.24

è®Ńç»ĥ

-0.23

 barriers

-0.23

çļ

-0.23

POSITIVE LOGITS

auth

0.27

 Georgia

0.26

Ñĵ

0.26

aight

0.25

akh

0.24

/auth

0.24

_CheckedChanged

0.24

ga

0.24

ass

0.24

æľ¯è¯Ń

0.23

Activations Density 0.002%

or

No Known Activations