Fri 13 Nov
LiveGet it free
ZOTPAPER

News without the noise

← All concepts

knowledge graph

A knowledge graph is a way of organizing information as a web of interconnected facts, where pieces of data (like a drug, a document, or a user) are linked by meaningful relationships (like "contains information about" or "interacts with"). Instead of storing data in isolated rows or files, it lets you navigate from one item to another through these connections, much like following links between web pages. This structure helps machines understand context and find relevant information that might be scattered across different sources.

The papers use knowledge graphs to bridge fragmented information and improve retrieval in complex domains. A common thread is combining the graph with large language models to ground reasoning in structured facts, leading to higher accuracy and safety in tasks like document retrieval, query generation, and clinical decision-making. The graph serves as a backbone that organizes messy data into queryable relationships, but its effectiveness depends on careful integration with AI agents, often reducing reliance on raw text similarity or unstructured memory.

Papers
14
Released code
0
First seen
Aug 2026
Latest
Nov 2026

10 papers in the last two months, against 4 in the two before.

Start here

Three papers to read first, in the order they came out
  1. 1Event graphs improve LLM stock forecasts across two major markets

    Temporally grounded, reliability-calibrated historical event transitions add incremental stock-forecasting signal when applied as a residual correction to a multi-view LLM forecast.

    Oct 2026
  2. 2AI can query undocumented company databases by learning from past queries

    Historical SQL logs can provide enough schema and relationship knowledge for entity-driven GraphRAG to generate more executable, higher-quality enterprise SQL than embedding-based retrieval.

    Oct 2026
  3. 3AI agents can pinpoint which code needs updating when designs change

    Representing design assets as a content-addressed Merkle directed acyclic graph turns large language model code evolution impact analysis into deterministic hash propagation.

    Oct 2026

How fast it is moving

JunJulAugSeptOctNov so far

Often studied with

agent memoryA repository where an agent stores past interactions and outcomes to inform future decisions.6 togetherlarge language modelA neural network trained on massive text corpora to generate coherent text and perform diverse language tasks.3 togetheragent safetyRisks from AI agents including timing errors, shutdown avoidance, collusion, overclaiming, and unsafe execution, mitigated by constrained decoding, safety harnesses, and swarm governance.2 together

The papers

Most central to this idea first, not most recent.

Concepts are extracted from each paper and reused across the corpus, so this page grows on its own as the desk reads.