INDEX

Explanations

extracting document substrings

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

mwhanna/qwen3-4b-transcoders/layer_25.safetensors

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

163,840

Data Type

float32

Hook Name

blocks.25.mlp.hook_in

Architecture

transcoder

Context Size

8,192

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Er

-0.27

 Rising

-0.27

 Mant

-0.27

ä¹Łå¥½

-0.25

Dem

-0.24

 convention

-0.24

 prayers

-0.24

Lim

-0.24

æļĿ

-0.24

çļĦæĥħ

-0.24

POSITIVE LOGITS

 reluct

0.35

eki

0.29

relude

0.27

athed

0.27

thr

0.27

pora

0.26

-io

0.25

 excuse

0.25

ernaut

0.25

éĽĨä½ĵç»ıæµİ

0.24

Activations Density 0.888%

extracting document substrings

No Comments

No Known Activations