INDEX

Explanations

Code and News

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

æ²»

-0.27

 Measurement

-0.27

::-

-0.26

ç¤¾ä¼ļ

-0.25

>(()

-0.25

Ĥ¨

-0.24

 Mutable

-0.24

Measurement

-0.23

 esac

-0.23

 Ð¶Ðµ

-0.23

POSITIVE LOGITS

tractive

0.28

adata

0.28

aggregate

0.28

 Arabic

0.28

anus

0.27

tee

0.27

è¾ľ

0.26

çĶľèľľ

0.26

çĪ±ä¸Ĭ

0.25

edge

0.25

Activations Density 6.141%

Code and News

No Known Activations