INDEX

Explanations

relationships or connections between different variables or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 pungkasnya

-0.36

 pasukan

-0.35

คร

-0.34

 vastaan

-0.34

 aikana

-0.34

 smě

-0.33

 tamam

-0.32

 procédé

-0.32

 dewasa

-0.31

 xong

-0.31

POSITIVE LOGITS

 relation

0.67

 correlation

0.65

 ligiloj

0.65

relation

0.64

 Relation

0.61

AddHtmlAttribute

0.60

 Correlation

0.60

 relaciona

0.59

ViewFeatures

0.59

 relacion

0.59

Activations Density 0.345%

relationships or connections between different variables or concepts

No Comments

No Known Activations