INDEX

Explanations

much

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

eux

-0.28

ÑģÑĥÐ´

-0.27

 them

-0.27

immers

-0.24

them

-0.24

Them

-0.24

äºĨè®¸å¤ļ

-0.24

_almost

-0.24

encies

-0.23

bus

-0.23

POSITIVE LOGITS

 alike

0.35

own

0.27

.ReadString

0.26

OfWork

0.26

 crow

0.26

 detail

0.25

Unused

0.25

Unnamed

0.24

acking

0.24

 composition

0.24

Activations Density 0.025%

much

No Comments

No Known Activations