INDEX

Explanations

technical and scientific data formatting elements

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

scientific notation or formulas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 autorytatywna

-0.85

 مرئيه

-0.83

تقاوى

-0.82

GEBURTSDATUM

-0.79

 Wicidata

-0.77

:✨

-0.75

WithIOException

-0.74

MLLoader

-0.73

Personendaten

-0.72

脚注の使い方

-0.70

POSITIVE LOGITS

 roue

0.33

Police

0.31

packung

0.31

 väg

0.31

 Again

0.30

getOption

0.29

 Leche

0.29

воду

0.28

Again

0.28

This

0.28

Activations Density 0.030%

technical and scientific data formatting elements

scientific notation or formulas

No Comments

No Known Activations