INDEX

Explanations

instances of numerical data and associated contextual phrases in scientific or academic writing

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

journal citations

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

AndEndTag

-0.60

InteropServices

-0.53

saraba

-0.50

Hentet

-0.49



-0.48

 Audiodateien

-0.47

GEBUR

-0.45

 مرئيه

-0.44

бята

-0.44

 defaultstate

-0.43

POSITIVE LOGITS

MigrationBuilder

0.61

AsUp

0.58

цездатний

0.55

 Италијани

0.49

LookAnd

0.45

 autorytatywna

0.45

:✨

0.45

 disambiguazione

0.44

 censiti

0.44

 PeEnEo

0.43

Activations Density 0.038%

instances of numerical data and associated contextual phrases in scientific or academic writing

journal citations

No Comments

No Known Activations