INDEX

Explanations

phrases that indicate numerical data and proportions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 تانيه

-0.60

 superiori

-0.58

 الحره

-0.57

 Zacks

-0.57

hiya

-0.55

 그것

-0.54

เฉ

-0.54

 critères

-0.54

 degradability

-0.53

 stethoscope

-0.53

POSITIVE LOGITS

 numberOf

1.14

numberOf

1.10

 antal

1.06

 aantal

1.05

 fewer

1.02

Aantal

1.02

NumberOf

1.01

 počet

0.96

 Anzahl

0.95

 Fewer

0.95

Activations Density 0.180%

phrases that indicate numerical data and proportions

No Comments

No Known Activations