INDEX

Explanations

references to authorship or attribution in content

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

copyright and author attributions

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

kuuta

-0.33

PINES

-0.32

 cuerda

-0.29

 fundos

-0.29

IntoConstraints

-0.28

✥

-0.26

실

-0.26

 tejidos

-0.25

ハワイ

-0.25

 てい

-0.25

POSITIVE LOGITS

 defaultstate

0.75

IBOutlet

0.75

GOTREF

0.74

OCCURRED

0.74

providedIn

0.72

parsedMessage

0.68

RefNanny

0.68

 nahilalakip

0.68

 snippetHide

0.67

httphttps

0.66

Activations Density 0.001%

references to authorship or attribution in content

copyright and author attributions

No Comments

No Known Activations