INDEX

Explanations

phrases or references associated with location and measurement

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

technical or foreign substrings

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

 الرياضيه

-0.56

 ModelExpression

-0.45

DeleteBehavior

-0.44

UserScript

-0.43

GHIJKLM

-0.43

 arşivlendi

-0.40

keit

-0.40

AccessorTable

-0.40

 متحده

-0.40

⟬

-0.40

POSITIVE LOGITS

 Administrativna

0.56

esModule

0.49

 snippetHide

0.41

HtmlAttribute

0.40

styleUrls

0.39

Πηγές

0.38

 surla

0.38

ConstraintMaker

0.37

+#+#

0.37

 الحره

0.36

Activations Density 0.009%

phrases or references associated with location and measurement

technical or foreign substrings

No Comments

No Known Activations