INDEX

Explanations

instances of coercion and related concepts 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 @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 légales

-0.34

Ram

-0.34

 ра

-0.34

 coastal

-0.33

 maux

-0.32

 Pauly

-0.32

 reparar

-0.32

Mal

-0.30

 cortical

-0.30

dó

-0.29

POSITIVE LOGITS

 renang

0.68

 NSCoder

0.67

 تضيفلها

0.65

parsedMessage

0.61

 Administrativna

0.60

queryInterface

0.59

IVEREF

0.57

 فريبيس

0.57

 AssemblyTitle

0.56

ConstraintMaker

0.56

Activations Density 0.631%

instances of coercion and related concepts in various contexts

No Known Activations