INDEX

Explanations

technical terms and jargon related to scientific research and methodologies

oai_token-act-pair · gpt-4o-mini Triggered by @bot

after bracketed text or symbols

np_acts-logits-general · gemini-2.0-flash

health, pharmaceutical, sensing

np_acts-logits-general · gemini-2.5-flash-lite

technical scientific-paper content, especially bioscience/chemistry jargon with Latin species names, chemical nomenclature, enzyme terms, and citation/figure/table formatting.

oai_token-act-pair · gpt-5 Triggered by @vetterc0

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Савезне

-1.69

Personendaten

-1.23

SourceChecksum

-1.16

:✨

-1.13

webElementXpaths

-1.09

 propOrder

-1.09

 Italijanski

-1.08

featureID

-1.06

 Италијани

-1.05

 мәкал

-1.05

POSITIVE LOGITS

 chien

0.23

 divino

0.23

 kirke

0.23

0.22

 dorada

0.22

 gobern

0.21

 trous

0.21

 trés

0.20

Activations Density 24.084%

technical terms and jargon related to scientific research and methodologies

after bracketed text or symbols

health, pharmaceutical, sensing

technical scientific-paper content, especially bioscience/chemistry jargon with Latin species names, chemical nomenclature, enzyme terms, and citation/figure/table formatting.

No Comments

No Known Activations