Author

Data Party THU

Official platform of Tsinghua Big Data Research Center, sharing the team's latest research, teaching updates, and big data news.

318

Articles

Likes

Views

Comments

Latest from Data Party THU

100 recent articles max

Data Party THU

Feb 25, 2026 · Artificial Intelligence

Why Multimodal LLMs Miss Tiny Objects—and How to Fix It

This article analyzes why multimodal large language models often fail to detect small objects, identifies three core bottlenecks, and presents a four‑tiered optimization roadmap—from zero‑cost inference tricks to data augmentation, model fine‑tuning, and engineering safeguards—backed by three real‑world case studies and actionable guidelines.

Data Augmentationinference optimizationmodel fine-tuning

0 likes · 20 min read

Why Multimodal LLMs Miss Tiny Objects—and How to Fix It

Data Party THU

Feb 24, 2026 · Artificial Intelligence

Why Long Contexts Undermine LLM Reliability: Hidden Risks of Personalization and Shared Sessions

The article analyzes how expanding the context window of large language models creates scarce attention, introduces unreproducible personalization, mixes intents in shared accounts, and leads to performance degradation, making debugging, testing, and reliable production deployment increasingly difficult.

AI reliabilitycontext managementpersonalization

0 likes · 11 min read

Why Long Contexts Undermine LLM Reliability: Hidden Risks of Personalization and Shared Sessions

Data Party THU

Feb 21, 2026 · Artificial Intelligence

Unlocking Compositional Generalization: Meta‑Learning Strategies for Neural Networks

This article examines how meta‑learning combined with compositionality enables neural networks to rapidly adapt to new tasks by formalizing hierarchical optimization, leveraging modular architectures with hypernetworks, and exploiting Transformer latent codes for effective compositional generalization.

Bilevel OptimizationTransformercompositional generalization

0 likes · 5 min read

Unlocking Compositional Generalization: Meta‑Learning Strategies for Neural Networks

Data Party THU

Feb 19, 2026 · Artificial Intelligence

How Data Priors and Scene Parameterization Boost 3D Indoor Reconstruction

This thesis investigates the two core challenges of data prior utilization and scene parameterization in multi‑view RGB‑based 3D indoor reconstruction, proposing novel representations and learning‑based methods to improve reconstruction quality, generalization, and applicability across AR, robotics, and autonomous navigation.

3D reconstructioncomputer visiondata priors

0 likes · 8 min read

How Data Priors and Scene Parameterization Boost 3D Indoor Reconstruction

Data Party THU

Feb 18, 2026 · Artificial Intelligence

Why Top AI Agents Fail in Real Work: Inside the Trainee‑Bench Benchmark

The article analyzes the gap between high benchmark scores and poor real‑world performance of AI agents, introduces the Trainee‑Bench workplace simulator, details its three evaluation dimensions, construction steps, and reveals that even state‑of‑the‑art models achieve low success rates, highlighting the need for autonomous learning and zero‑hand‑over.

AI AgentsTrainee-Benchcontinuous learning

0 likes · 11 min read

Why Top AI Agents Fail in Real Work: Inside the Trainee‑Bench Benchmark

Data Party THU

Feb 15, 2026 · Artificial Intelligence

Why Retrieval‑Augmented Generation Is Still Fragile: Boosting Generalization and Evidence‑Based Answers

Although modern information access is faster than ever, retrieval‑augmented generation systems remain vulnerable, especially when faced with distribution shifts, making it crucial to improve both retriever generalization across domains and languages and ensure generators produce evidence‑grounded responses or refuse when evidence is lacking.

AI robustnessRAGevidence grounding

0 likes · 3 min read

Why Retrieval‑Augmented Generation Is Still Fragile: Boosting Generalization and Evidence‑Based Answers

Data Party THU

Feb 15, 2026 · Artificial Intelligence

Why FireRed-Image-Edit Is the New Powerhouse in AI Image Editing

FireRed-Image-Edit, the latest open‑source image‑editing model from the Xiaohongshu Super Intelligence team, outperforms existing benchmarks with superior instruction understanding, ID preservation and efficient architecture, thanks to its RedEdit Bench evaluation suite, a three‑stage training pipeline and a scalable data‑engine.

AI Image EditingFireRed-Image-EditOpen Source

0 likes · 8 min read

Why FireRed-Image-Edit Is the New Powerhouse in AI Image Editing

Data Party THU

Feb 14, 2026 · Industry Insights

From Student to NVIDIA: 11 Years of AI Research Lessons

This reflective essay chronicles Zhaocheng Zhu’s eleven‑year journey from undergraduate AI curiosity through doctoral struggles, industry internships, and finally landing a research role at NVIDIA, offering candid insights on publishing, engineering, mentorship, and the evolving realities of academic and corporate AI work.

AIAdviceIndustry Transition

0 likes · 21 min read

From Student to NVIDIA: 11 Years of AI Research Lessons

Data Party THU

Feb 12, 2026 · Artificial Intelligence

10 Expert Tips to Boost Your Claude Code Productivity

This guide translates and refines the Claude Code founder’s ten practical usage tips, covering parallel worktree sessions, plan‑mode workflows, CLAUDE.md maintenance, custom skills, automated bug fixing, prompt strategies, terminal setup, sub‑agents, data analysis, and learning techniques to help developers get the most out of the AI coding assistant.

AI coding assistantClaude CodeData Analysis

0 likes · 10 min read

10 Expert Tips to Boost Your Claude Code Productivity

Data Party THU

Feb 9, 2026 · Artificial Intelligence

Aligning Collaborative Filtering with LLM Token Generation: The TCA4Rec Breakthrough

This paper introduces the TCA4Rec framework that directly aligns item‑level collaborative‑filtering preferences with token‑level objectives of large language models, presenting novel modules, extensive experiments, and analysis that demonstrate significant performance gains in generative recommendation tasks.

Generative RecommendationLLMTCA4Rec

0 likes · 9 min read

Aligning Collaborative Filtering with LLM Token Generation: The TCA4Rec Breakthrough