INDEX

Explanations

technical/code contexts

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_30.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.30.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 face

-0.32

 CAPITAL

-0.29

pheric

-0.28

-face

-0.28

 capital

-0.28

 Capital

-0.28

èĦ¸

-0.27

face

-0.27

Vel

-0.27

 rhet

-0.27

POSITIVE LOGITS

EPS

0.31

çº¢åĪ©

0.31

upo

0.29

opol

0.28

lining

0.27

çķĻåŃĺ

0.26

apo

0.26

çĽĳçĿ£æ£ĢæŁ¥

0.26

ç®¡çĲĨå±Ģ

0.26

 getElement

0.25

Activations Density 0.886%

technical/code contexts

No Comments

No Known Activations