INDEX

Explanations

references to experimental methodology and statistical terms

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

prepositional phrases

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

 للاسماء

-0.82

 Perſ

-0.74

évaluateur

-0.74

GOTREF

-0.70

 kaarangay

-0.70

verwijspagina

-0.68

 Reſ

-0.63

 ſche

-0.62

:✨

-0.60

INTENANCE

-0.60

POSITIVE LOGITS

MathML

0.30

 penanganan

0.30

enumi

0.27

\{\\

0.26

 møte

0.26

 néz

0.25

 disponibilités

0.24

 czasu

0.24

TabIndex

0.24

 tertentu

0.24

Activations Density 0.192%

references to experimental methodology and statistical terms

prepositional phrases

No Comments

No Known Activations