skills.rest
ElemontCapital
ElemontCapitalCOMMUNITY
Explore all skills by this author →
𝕏@ElemontGitHub

x-data-signals

Decode SimClusters, RealGraph, TweepCred, and TwHIN signals for candidate ranking pipelines.

2|Updated Jan 21, 2026
Search Tags:#embeddings#retrieval#signals#ranking#knowledge-graph#simclusters#twhin
One-click install
npx skills add https://github.com/ElemontCapital/x-algorithm-skills --skill x-data-signals
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: x-data-signals
Source: https://github.com/ElemontCapital/x-algorithm-skills/tree/main/plugins/x-algorithm/skills/x-data-signals
Command: npx skills add https://github.com/ElemontCapital/x-algorithm-skills --skill x-data-signals

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Decode core signals powering candidate generation and ranking.

Core Features & Use Cases

  • Signal decoding: exposes SimClusters, RealGraph, TweepCred, and TwHIN signals for retrieval and ranking.
  • Candidate generation: improves early-stage candidate selection by leveraging embedding-based similarities.
  • Use Case: map user embeddings to candidate content to surface relevant posts in the feed.

Quick Start

Ask the system to decode the core signals and generate top candidate embeddings for a given user context.

Frequently Asked Questions about x-data-signals

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What are SimClusters and TwHIN embeddings used for in candidate retrieval?▼

SimClusters and TwHIN embeddings map user interests to candidate content for retrieval. They enable dot-product scoring to measure embedding similarities and surface relevant posts in feeds.

How do I use embeddings for candidate generation in ranking pipelines?▼

Embeddings drive candidate generation by calculating dot-product scores between user and content vectors. You apply these scores in retrieval and ranking pipelines to filter candidates before the Heavy Ranker.

Can I use RealGraph and TweepCred signals for lightweight candidate filtering?▼

Yes, RealGraph and TweepCred signals support lightweight candidate filtering. They expose graph connectivity and user credibility metrics to refine early-stage candidate selection.

What is the best way to decode core signals for feed experiments?▼

Decoding core signals involves accessing SimClusters, RealGraph, TweepCred, and TwHIN data. You apply these signals in retrieval pipelines to generate top candidate embeddings for feed experiments and validation tasks.

Does this approach support dot-product scoring in the Heavy Ranker?▼

Yes, dot-product scoring is supported for the Heavy Ranker. You compute similarities between user and candidate embeddings to provide lightweight candidate filtering and scoring inputs.

When should I not use embedding-based similarities for candidate selection?▼

Embedding-based similarities may not suit contexts lacking sufficient user interaction data. If user embeddings are sparse, dot-product scoring may fail to surface relevant candidates effectively.

Related Skills

Explore More Data & Analytics Skills→
cutec-chriscutec-chris
1

memory

Index daily chat logs and enable semantic search across past conversations.

Community
Intermediate
ElemontCapitalElemontCapital
2

x-ranking-engine

Explain Heavy Ranker scoring with calibrated probabilities and action weights.

Community
Advanced
vignesh2027vignesh2027
6

learning-coach

Designs adaptive curricula and assessments with Bloom's taxonomy alignment.

Community
Advanced
hiyenwonghiyenwong
2

bridge-neurodivergence-detection

Quantify neural divergence from typical developmental trajectories using graph-regularized modeling and deep generative networks.

Community
Advanced