INDEX

Explanations

terms related to measurement and evaluation in various contexts

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ivably

-0.40

}{*}{}

-0.37

CEAN

-0.35

Diverse

-0.34

 diverse

-0.33

 denn

-0.30

ieni

-0.30

 Autorizaciones

-0.30

antly

-0.29

何况

-0.29

POSITIVE LOGITS

الحياه

0.85

 disambiguazione

0.78

تقاوى

0.78



0.77

hyrchwyd

0.71

Diwedd

0.67

sidemargin

0.66

 lenker

0.66

TemporalType

0.64

GeoNames

0.64

Activations Density 0.129%

terms related to measurement and evaluation in various contexts

No Comments

No Known Activations