INDEX

Explanations

repeated or recurring concepts and agreements in a discussion or narrative

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

QL

-0.46

ACI

-0.45

SMP

-0.44

Arma

-0.43

 Arma

-0.43

 ples

-0.43

infor

-0.43

Ano

-0.42

)‏

-0.42

 Vira

-0.42

POSITIVE LOGITS

 Wikiseite

0.78

 dezelve

0.77

 zijne

0.77

 disambiguazione

0.71

 Offisielt

0.71

 ainfi

0.70

 feroit

0.69

 تضيفلها

0.68

 mijne

0.68

 enfans

0.67

Activations Density 0.240%

repeated or recurring concepts and agreements in a discussion or narrative

No Comments

No Known Activations