GAIA

GAIA dataset GAIA is a benchmark which aims at evaluating next-generation LLMs (LLMs with augmented capabilities due to added tooling, efficient prompting, access to search, etc). We added gating to prevent bots from scraping the dataset. Please do not reshare the validation or test set in a crawlable format. Data and

Papers using GAIA (16)

PACE: A Proxy for Agentic Capability Evaluation2026SPORK: Self-Speculative Forking to Accelerate Agentic LLM Inference2026Towards Direct Latent-Space Synthesis for Parallel Branches in LLM-Agent Workflows2026EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments2026MemoBrain: Executive Memory as an Agentic Brain for Reasoning2026Learning Agentic Policy from Action Guidance2026Beyond Reward Engineering: A Data Recipe for Long-Context Reinforcement Learning2026MiroFlow: Towards High-Performance and Robust Open-Source Agent Framework for General Deep Research Tasks2026Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution2025Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPO2025DoVer: Intervention-Driven Auto Debugging for LLM Multi-Agent Systems2025MetaChain: A Fully-Automated and Zero-Code Framework for LLM Agents2025AgentFly: Fine-tuning LLM Agents without Fine-tuning LLMs2025Where LLM Agents Fail and How They can Learn From Failures2025WebLeaper: Empowering Efficiency and Efficacy in WebAgent via Enabling Info-Rich Seeking2025Youtu-Agent: Scaling Agent Productivity with Automated Generation and Hybrid Policy Optimization2025