INDEX

Explanations

words and phrases that suggest characteristics or concepts of databases, living beings, and roles within narratives

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

 prédio

-0.37

 Fahrt

-0.37

 Aspekte

-0.34

 orilla

-0.33

sivité

-0.33

 origini

-0.33

 istrinya

-0.32

ญิง

-0.32

 Finanzierung

-0.32

 Formación

-0.32

POSITIVE LOGITS

 Савезне

0.64

divarius

0.56

 censiti

0.55

ollectionView

0.52

千里

0.50

phro

0.49

petual

0.49

 diplo

0.49

 virtual

0.48

ジアム

0.48

Activations Density 0.068%

words and phrases that suggest characteristics or concepts of databases, living beings, and roles within narratives

No Comments

No Known Activations