INDEX

Explanations

specific scientific and medical terminology, particularly relating to study methodologies and statistical analysis

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

descriptive words followed by specific terms

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

Билгалдахарш

-0.55

LookAnd

-0.50

 ainfi

-0.48

antaranya

-0.46

 utafitiHapana

-0.46

 orejas

-0.46

 invitaciones

-0.45

 Económica

-0.45

 invitados

-0.45

 placas

-0.44

POSITIVE LOGITS

CloseOperation

0.43

esgue

0.42

DeleteBehavior

0.42

يكب

0.42

oredCriteria

0.40

DockStyle

0.39

uple

0.38

LLocation

0.38

TargetException

0.38

parsedMessage

0.37

Activations Density 0.145%

specific scientific and medical terminology, particularly relating to study methodologies and statistical analysis

descriptive words followed by specific terms

No Comments

No Known Activations