MaDI-Bench supplies the first end-to-end benchmark tasks for full relational data integration pipelines across domains plus a variant-generation method to slow saturation.
hub
A survey of data agents: Emerging paradigm or overstated hype?CoRR, abs/2510.23587
15 Pith papers cite this work. Polarity classification is still indexing.
15
Pith papers citing it
hub tools
citation-role summary
baseline 1
citation-polarity summary
years
2026 15roles
baseline 1polarities
baseline 1representative citing papers
The paper introduces a layered vulnerability framework and attack taxonomy for LLM-driven data agents and demonstrates attacks on four open-source and two production systems.
PrepBench is a benchmark showing that state-of-the-art LLMs still struggle with natural-language-driven data preparation involving disambiguation, code generation, and workflow translation.
DataPRM is an environment-aware generative process reward model that improves LLM data analysis agents by 7-11% on benchmarks via active verification and reflection-aware ternary rewards.
ROSE is an intent-centered NL2SQL metric using an adversarial Prover-Refuter cascade that achieves higher human-expert agreement than prior metrics on a new validation set.
SQLConductor uses Search-to-Policy Learning with MCTS, stability-weighted SFT, and curriculum RL to train a compact policy for adaptive step-wise Text-to-SQL orchestration, reporting 73.2% EX on BIRD-Dev.
EvoDS adds autonomous skill acquisition via synthesis-validation-reuse and adaptive context compression via learned control within a two-stage multi-agent RL scheme, claiming 28.9% average gains over prior agents on four benchmarks plus elimination of out-of-token failures.
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
DataMagic generates narrative data videos from tabular data and queries via DVSpec declarative bindings and a Generate-then-Orchestrate multi-agent pipeline.
DataCOPE uses verifier-guided contrastive distillation from agent trajectories to discover skills, yielding average gains of 9.71% on report-style and 32.30% on reasoning-style data analysis tasks across four model settings.
This survey defines execution provenance as a typed graph of agent execution and evidence tracing as its projection onto evidence-support relations, then reviews methods, taxonomy, benchmarks, and challenges for auditable LLM agents.
CP-Agent improves LLM competitive programming performance via calibrated feedback mechanisms that target false-admission risk, evidence against bad programs, and success hazard.
Generalizable agents require environment scaling via diverse executable rule-sets, distinguished from trajectory and task scaling in a new taxonomy.
Describes a multi-agent system for proactive insight discovery in real-time streaming data via LLMs and streaming platforms.
citing papers explorer
-
MaDI-Bench: An End-to-End Data Integration Benchmark
MaDI-Bench supplies the first end-to-end benchmark tasks for full relational data integration pipelines across domains plus a variant-generation method to slow saturation.
-
Data Agents Under Attack: Vulnerabilities in LLM-Driven Analytical Systems
The paper introduces a layered vulnerability framework and attack taxonomy for LLM-driven data agents and demonstrates attacks on four open-source and two production systems.
-
PrepBench: How Far Are We from Natural-Language-Driven Data Preparation?
PrepBench is a benchmark showing that state-of-the-art LLMs still struggle with natural-language-driven data preparation involving disambiguation, code generation, and workflow translation.
-
Rewarding the Scientific Process: Process-Level Reward Modeling for Agentic Data Analysis
DataPRM is an environment-aware generative process reward model that improves LLM data analysis agents by 7-11% on benchmarks via active verification and reflection-aware ternary rewards.
-
ROSE: An Intent-Centered Evaluation Metric for NL2SQL
ROSE is an intent-centered NL2SQL metric using an adversarial Prover-Refuter cascade that achieves higher human-expert agreement than prior metrics on a new validation set.
-
SQLConductor: Search-to-Policy Learning for Step-wise Text-to-SQL Orchestration
SQLConductor uses Search-to-Policy Learning with MCTS, stability-weighted SFT, and curriculum RL to train a compact policy for adaptive step-wise Text-to-SQL orchestration, reporting 73.2% EX on BIRD-Dev.
-
EvoDS: Self-Evolving Autonomous Data Science Agent with Skill Learning and Context Management
EvoDS adds autonomous skill acquisition via synthesis-validation-reuse and adaptive context compression via learned control within a two-stage multi-agent RL scheme, claiming 28.9% average gains over prior agents on four benchmarks plus elimination of out-of-token failures.
-
Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents
A data-centric survey finds that only information-flow control covers compositional and cross-session leakage in LLM agents and that no single benchmark tests an agent across all its data surfaces under one policy.
-
DataMagic: Transforming Tabular Data into Data Insight Video
DataMagic generates narrative data videos from tabular data and queries via DVSpec declarative bindings and a Generate-then-Orchestrate multi-agent pipeline.
-
Unsupervised Skill Discovery for Agentic Data Analysis
DataCOPE uses verifier-guided contrastive distillation from agent trajectories to discover skills, yielding average gains of 9.71% on report-style and 32.30% on reasoning-style data analysis tasks across four model settings.
-
From Agent Traces to Trust: A Survey of Evidence Tracing and Execution Provenance in LLM Agents
This survey defines execution provenance as a typed graph of agent execution and evidence tracing as its projection onto evidence-support relations, then reviews methods, taxonomy, benchmarks, and challenges for auditable LLM agents.
-
CP-Agent: A Calibrated Risk-Controlled Agent for Feedback-Driven Competitive Programming
CP-Agent improves LLM competitive programming performance via calibrated feedback mechanisms that target false-admission risk, evidence against bad programs, and success hazard.
-
Scalable Environments Drive Generalizable Agents
Generalizable agents require environment scaling via diverse executable rule-sets, distinguished from trajectory and task scaling in a new taxonomy.
-
Discovery Agents for Real-Time Analytics: Toward Proactive Insight Systems
Describes a multi-agent system for proactive insight discovery in real-time streaming data via LLMs and streaming platforms.
- An Agentic Approach to Metadata Reasoning