INDEX

Explanations

references to numerical values or indicators

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

history of table of contents

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

 ब्रेकडाउन

-0.65

httphttps

-0.60

Manbalar

-0.50

fitrión

-0.46

évaluateur

-0.45

 Laub

-0.44

Билгалдахарш

-0.44

脚注の使い方

-0.44

ńskiej

-0.44

 برانيه

-0.43

POSITIVE LOGITS

InjectAttribute

0.41

 CURIAM

0.40

 inversores

0.39

 förr

0.38

XmlAccessorType

0.37

verhalten

0.35

digd

0.35

Filmographie

0.34

 ویکی‌پدیا

0.34

DataPropertyName

0.33

Activations Density 0.000%

references to numerical values or indicators

history of table of contents

No Comments

No Known Activations