INDEX

Explanations

numerical values and dates in the text

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

date and citation markers

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

 INTERESAR

-0.44

 <<<<<<<<<<<<<<

-0.44

ⓧ

-0.42

SharedCtor

-0.41

تقاوى

-0.41

 linkovi

-0.40

delwed

-0.40

bosky

-0.39

WebElementEntity

-0.39

 detailsNormal

-0.39

POSITIVE LOGITS

PRWEB

0.52



0.48

 initComponents

0.45

autoreleasepool

0.44

 Peer

0.43

 noDo

0.43

 EconPapers

0.41

Saiba

0.40

caneca

0.39

 anuncia

0.39

Activations Density 0.001%

numerical values and dates in the text

date and citation markers

No Comments

No Known Activations