INDEX

Explanations

references to specific coding or programming syntax and terminologies

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

latex formatting suppressors

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

HtmlAttribute

-0.47

ukone

-0.39

WebServlet

-0.37

fromnode

-0.36

hassee

-0.34

 giri

-0.33

MouseListener

-0.33

PositiveButton

-0.33

 arrivent

-0.32

ワシ

-0.32

POSITIVE LOGITS

 noqa

4.94

OGND

0.77

 kasarigan

0.76

 nahilalakip

0.73

 <<<<<<<<<<<<<<

0.69

uxxxx

0.63

httphttps

0.62

 autorytatywna

0.60

ValueStyle

0.58

 فريبيس

0.57

Activations Density 0.001%

references to specific coding or programming syntax and terminologies

latex formatting suppressors

No Comments

No Known Activations