arxiv:2610.26770
🔄 In a Training Loop
Honglin Lin
LHL3341
AI & ML interests
None yet
Recent Activity
authored a paper 3 days ago
HarnessPro: Recursive Harness Self-Improvement with Pareto Optimization for Performance-Cost Trade-offs authored a paper 3 days ago
ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL authored a paper 3 days ago
Tracing the Roots: A Multi-Agent Framework for Uncovering Data Lineage in Post-Training LLMsOrganizations
arxiv:2608.28476
arxiv:2604.10480
arxiv:2604.06079
arxiv:2603.07223
arxiv:2601.21821
arxiv:2601.13606
arxiv:2601.17027
arxiv:2512.14051
arxiv:2510.04081
arxiv:2509.21070
arxiv:2507.17512
arxiv:2504.19093
arxiv:2504.12322
arxiv:2503.17439
arxiv:2503.16212
arxiv:2503.14891
arxiv:2412.17007
arxiv:2410.09732
