INDEX

Explanations

Proper names

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

isman

-0.27

sv

-0.27

parated

-0.27

äº²çĪ±

-0.26

@author

-0.26

éĹª

-0.26

witter

-0.24

 sight

-0.24

ç¬¦

-0.24

-dot

-0.23

POSITIVE LOGITS

io

0.29

iano

0.27

 UIGraphics

0.26

reno

0.26

Um

0.26

icum

0.26

ieri

0.25

ierz

0.25

lus

0.25

gil

0.24

Activations Density 0.004%

Proper names

No Comments

No Known Activations