Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents
Paper • 2610.20344 • Published • 97
AI & ML interests
None defined yet.
Recent Activity
View all activity
Papers
Doc2Harness: Benchmarking LLMs on Long-Horizon Harness Engineering of Coding Agents
SkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles
YuyaoGe
authored a
paper about 19 hours ago
YuyaoGe
submitted a
paper to Daily Papers 1 day ago
YuyaoGe
authored 4
papers 7 days ago
SkillForge: Co-Evolving Skills and Agents via Dynamic Skill Lifecycles
Paper • 2610.09832 • Published • 14
Kimi K2.5: Visual Agentic Intelligence
Paper • 2602.02276 • Published • 279
Gated Differentiable Working Memory for Long-Context Language Modeling
Paper • 2601.12906 • Published
Reward and Guidance through Rubrics: Promoting Exploration to Improve Multi-Domain Reasoning
Paper • 2511.12344 • Published • 1
YuyaoGe
submitted a
paper to Daily Papers 8 days ago
Post
101
🤖 We train LLMs — but how do LLMs reshape AI research itself?
Sharing our new survey, AI4AIR: A Comprehensive Survey on Large Language Models for AI Research 📚
Rather than treating LLMs as outside helpers for writing or literature review, we map how they enter the ML pipeline itself — through a 2D taxonomy (research domain × pipeline stage) and 5 recurring roles:
🏷️ Annotator · 🧪 Synthesizer · ⚙️ Optimizer · 📊 Evaluator · 🎛️ Orchestrator
Our key takeaway: how deeply an LLM can empower a stage is bounded by its validation cost — which is exactly why data annotation & evaluation are already mature, while model design & full-loop orchestration remain hard.
📄 Paper (PDF): https://github.com/ICT-FinD-Lab/Awesome-LLMs-for-AI-Research/blob/master/PDF/AI4AIR_Survey_v260601.pdf
🌐 Project: https://ict-find-lab.github.io/Awesome-LLMs-for-AI-Research
⭐ GitHub: https://github.com/ICT-FinD-Lab/Awesome-LLMs-for-AI-Research
Stars & feedback warmly welcome 🙌
Sharing our new survey, AI4AIR: A Comprehensive Survey on Large Language Models for AI Research 📚
Rather than treating LLMs as outside helpers for writing or literature review, we map how they enter the ML pipeline itself — through a 2D taxonomy (research domain × pipeline stage) and 5 recurring roles:
🏷️ Annotator · 🧪 Synthesizer · ⚙️ Optimizer · 📊 Evaluator · 🎛️ Orchestrator
Our key takeaway: how deeply an LLM can empower a stage is bounded by its validation cost — which is exactly why data annotation & evaluation are already mature, while model design & full-loop orchestration remain hard.
📄 Paper (PDF): https://github.com/ICT-FinD-Lab/Awesome-LLMs-for-AI-Research/blob/master/PDF/AI4AIR_Survey_v260601.pdf
🌐 Project: https://ict-find-lab.github.io/Awesome-LLMs-for-AI-Research
⭐ GitHub: https://github.com/ICT-FinD-Lab/Awesome-LLMs-for-AI-Research
Stars & feedback warmly welcome 🙌
CapitalLiu
submitted a
paper to Daily Papers 5 months ago
ZhenlongYuan
submitted a
paper to Daily Papers 5 months ago
YuyaoGe
updated a
Space 6 months ago
YuyaoGe
published a
Space 6 months ago
YuyaoGe
authored a
paper 7 months ago
YuyaoGe
submitted a
paper to Daily Papers 7 months ago
THEATLAS
submitted a
paper to Daily Papers 9 months ago
Shiyunee
authored a
paper 10 months ago
Shiyunee
authored 2
papers 12 months ago
YuyaoGe
authored 2
papers about 1 year ago